Genetically diverse uropathogenic Escherichia coli adopt a common transcriptional program in patients with UTIs.

Uropathogenic Escherichia coli (UPEC) is the major causative agent of uncomplicated urinary tract infections (UTIs). A common virulence genotype of UPEC strains responsible for UTIs is yet to be defined, due to the large variation of virulence factors observed in UPEC strains. We hypothesized that studying UPEC functional responses in patients might reveal universal UPEC features that enable pathogenesis. Here we identify a transcriptional program shared by genetically diverse UPEC strains isolated from 14 patients during uncomplicated UTIs. Strikingly, this in vivo gene expression program is marked by upregulation of translational machinery, providing a mechanism for the rapid growth within the host. Our analysis indicates that switching to a more specialized catabolism and scavenging lifestyle in the host allows for the increased translational output. Our study identifies a common transcriptional program underlying UTIs and illuminates the molecular underpinnings that likely facilitate the fast growth rate of UPEC in infected patients.

Introduction

Urinary tract infections (UTIs) are among the most common bacterial infections in humans, affecting 150 million people each year worldwide (Flores-Mireles et al., 2015). A high incidence of recurrence and frequent progression to chronic condition exacerbates the negative impact of UTIs on patients’ quality of life and healthcare cost (Foxman, 2010). Despite the magnitude of the problem, treatment remains limited by a strain’s susceptibility to available antibiotics, which are often ineffectual (Albert et al., 2004; Nickel, 2005; Sihra et al., 2018).

The major causative agent of uncomplicated UTIs is Uropathogenic Escherichia coli (UPEC), which is responsible for upwards of 70% of all cases (Flores-Mireles et al., 2015). The majority of our insights into UPEC pathogenesis have been obtained through in vitro assays, cell culture systems, and animal models (Alteri et al., 2009; Alteri and Mobley, 2015; Sivick and Mobley, 2010; Subashchandrabose and Mobley, 2015). While these studies have identified virulence and fitness factors that are important for UPEC infection, how these studies translate to human infection is not clear. As a result, we do not yet have a complete understanding of UPEC physiology in the human urinary tract. Moreover, the genetic heterogeneity of UPEC isolates, which carry diverse and functionally redundant virulence systems including iron acquisition, adherence, and toxins, further complicates our understanding of uropathogenesis (Johnson et al., 1998; Johnson et al., 2001; Köhler and Dobrindt, 2011; Schreiber et al., 2017; Takahashi et al., 2006). The different constellations of virulence factors and diverse genetic backgrounds raise the question of whether different UPEC strains vary in their strategies for pathogenesis.

Since defining conserved UPEC characteristics have proven elusive to comparative genomics strategies, we hypothesized that comparing functional responses in the context of the host may uncover disease-defining features. To that end, we examined UPEC gene expression directly from 14 patients with documented significant bacteriuria and presenting with uncomplicated UTI and compared it to the gene expression of the identical strains cultured to mid-exponential stage in filter-sterilized pooled human urine. Despite the genetic diversity of the pathogen and the human hosts, we identified a remarkably conserved gene expression program that is specific to human infection and strongly supports previous findings of extremely rapid UPEC growth rate during UTI (Bielecki et al., 2014; Burnham et al., 2018; Forsyth et al., 2018). Importantly, we show that this transcriptional program is recapitulated in the mouse model of infection and propose a mechanism by which the fast growth rate can be achieved. Based on extensive analysis, we propose a model where UPEC shut down all non-essential metabolic processes and commit all available resources to rapid growth during human UTI. Critically, our discovery of a common transcriptional program of UPEC in patients significantly expands our understanding of bacterial adaptation to the human host and provides a platform to design universal therapeutic strategies.

Results

Study design

To better understand UPEC functional responses to the human host, we isolated and sequenced RNA from the urine (stabilized immediately after collection) from fourteen otherwise healthy women diagnosed with UPEC-associated urinary tract infection. To identify infection-specific responses, we cultured the same fourteen UPEC isolates in vitro in filter-sterilized human urine (mid-exponential phase, 2 hr time point in Figure 1—figure supplement 1), and isolated and sequenced RNA from these cultures (study design and quality control is described in detail in Methods section). Phylogenetic analysis showed a high degree of genetic diversity, as we identified strains belonging to three distinct phylogroups, 13 different sequence types, and 13 distinct serogroups (Figure 1—figure supplement 2, Table 1, Table 2). The majority of UPEC isolates (10 of 14) belonged to the B2 phylogroup, which is consistent with previously published studies (Foxman, 2010; Schreiber et al., 2017). Although the majority (10 of 14) of patients had a previous history of UTIs, we found no relationship between patients’ previous UTI history and bacterial genotype (Figure 1—figure supplement 2). Moreover, the 14 clinical isolates showed a wide array of antibiotic resistance phenotypes (Figure 1—figure supplement 2).

Figure 1—figure supplement 1.

Growth curves for 14 clinical UPEC strains cultured in LB or filter-sterilized urine.

Figure 1—figure supplement 2.

Phylogenetic tree reconstruction of 14 clinical UPEC strains isolated in this study.

Antibiotic resistance profiles are indicated by filled in black circles (as determined by VITEK2 system (BioMerieux).) Patients with recurrent UTIs are indicated by filled in black square. MG1655, EC958, UTI89 and CFT073 are included for reference.

Table 1.

Sequence type for 14 clinical UPEC isolates

Strain	Sequence type	Adk	fumC	gyrB	Icd	Mdh	purA	recA
HM01	69	21	35	27	6	5	5	4
HM03	101	43	41	15	18	11	7	6
HM06	131	53	40	47	13	36	28	29
HM07	641*	9	6	33*	131	24	8	7
HM14	Novel	6	4	4	16	24	13	14
HM17	73	36	24	9	13	17	11	25
HM43	Novel*	40*	14	19	36	17	10	203
HM54	404*	14*	14	10	14	17	7	74
HM56	538	13	40	19	13	36	28	30
HM57	73	36	24	9	13	17	11	25
HM60	648	92	4	87	96	70	58	2
HM66	80	13	24	19	14	23	1	10
HM68	998	13	52	156	14	17	25	17
HM86	127	13	14	19	36	23	11	10

Table 2.

In silico determined serotypes for 14 clinical UPEC strains

Strain	H_type	O_type
HM01	H4	O25
HM03	H21	NA
HM06	H4	O25
HM07	H45	O45
HM14	H10	O8
HM17	H1	O6
HM43	H23	NA
HM54	H5	O75
HM56	H4	O13/O135
HM57	H1	O2/O50
HM60	H10	O102
HM66	H7	O7
HM68	H6	O2/O50
HM86	H31	O6

Virulence factor expression is observed both during urine culture and human infection

We first assessed the virulence genotype of the fourteen UPEC strains by looking at the presence or absence of a comprehensive list of known virulence factors, including adhesins, toxins, iron acquisition proteins, and flagella (Johnson et al., 2001; Johnson and Stell, 2000; Köhler and Dobrindt, 2011; Schreiber et al., 2017; Subashchandrabose and Mobley, 2015; Tarchouna et al., 2013) (Figure 1A). As previously reported (Schreiber et al., 2017), B1 strains appear to carry fewer virulence factors overall when compared to B2 strains, suggesting that UTIs can be established by UPEC strains with vastly diverse virulence genotypes. We then compared the levels of gene expression of these virulence factors following culture in filter-sterilized urine (Figure 1B, Figure 1—figure supplement 3) to that during infection. As expected, we detected expression of genes involved in iron acquisition during both in vitro urine culture and human UTI (Figure 1B). However, we also observed high strain-to-strain variability in gene expression, especially for hma, iutA, iucC and fyuA, which is consistent with previous reports (Subashchandrabose et al., 2014).

Figure 1.

Clinical UPEC isolates carry a highly variable set of virulence factors.

Phenotypic and genotypic information about the strains can be found in Figure 1—figure supplement 1, Figure 1—figure supplement 2, Table 1, and Table 2. (A) Clinical UPEC isolates were examined for presence of 40 virulence factors. Virulence factors were identified based on homology using BLAST searches (≥80% identity,≥90% coverage). The heatmap shows presence (black) or absence (white) of virulence factors across 14 UPEC strains. Hierarchical clustering based on presence/absence of virulence factors shows separate clustering of B1 isolates. (B) Log2 TPM for iron acquisition genes (top panel) and adhesins (bottom panel) in urine and patient samples. Gene expression of other virulence factors is shown in Figure 1—figure supplement 3. Correlations of virulence factor expression among in vitro and patient samples is shown in Figure 1—figure supplement 4. (C) Log2 TPM of fim (top panel) and flg (bottom panel) operons across the 14 UPEC strains during in vitro urine culture and human UTI.

Antibiotic resistance profiles are indicated by filled in black circles (as determined by VITEK2 system (BioMerieux).) Patients with recurrent UTIs are indicated by filled in black square. MG1655, EC958, UTI89 and CFT073 are included for reference.

Gene expression of a panel of genes was measured for HM86 (n = 3), and HM56 (n = 2) after 5 hr culture in filter-sterilized urine. After isolation RNA samples were either treated with MICROBEnrich or left untreated. Gene expression for each gene was measured by qRT-PCR. ΔCt between the gene of interest and gapA is shown.

Figure 1—figure supplement 3.

Expression of virulence factor genes in urine and patient samples.

Figure 1—figure supplement 4.

Correlations among in vitro and patient samples measured by Pearson correlation coefficient of normalized gene expression of 40 virulence factors plotted according to hierarchical clustering of samples.

Most of the adhesin genes were expressed at very low levels both during in vitro culture and infection, with the exception of fim genes (Figure 1B). Interestingly, we observed high variability in fim and flg operon expression between patients (Figure 1C). In the majority of the cases, we detected high levels of fim operon expression (9/14) and low levels of flg operon expression (12/14). However, in the sample collected from patient HM07, we observed high levels of both fim and flg expression, potentially indicating a mixed population of both motile and adherent bacteria present in the sample. Overall, the variability in the expression of adhesin and motility machinery might suggest different stages of infection.

Other virulence factors examined were expressed at either similar or lower levels during human UTI compared to in vitro urine cultures (Figure 1—figure supplement 3). Notably, virulence factor carriage varies greatly between UPEC strains and we did not discern any infection-specific gene expression among the virulence factors we examined (Figure 1—figure supplement 4).

The UPEC core genome exhibits a common gene expression program during clinical infection

Since patient samples contained fewer bacterial reads compared to in vitro controls, we first performed a rigorous quality assurance analysis, which indicated that we possessed sufficient sequencing depth for downstream analyses (Table 3, Table 4, Figure 2—figure supplement 1, Figure 2—figure supplement 2, see Materials and methods for details). Next, to perform a comprehensive comparison of gene expression between the different clinical UPEC strains, we identified a set of 2653 genes present in all 14 UPEC strains in this study as well as the reference E. coli MG1655 strain (hereafter referred to as the core genome). We then compared the gene expression correlation of the core genome to that of the accessory genome (i.e., 2219 genes that were present in at least two but not all of the clinical UPEC strains) for all 14 isolates cultured in vitro in filter-sterilized urine. As expected for bacterial strains cultured under identical conditions, we saw high correlation of gene expression between any two isolates cultured in vitro irrespective of whether these genes were part of the core or accessory genome (Figure 2A). Remarkably, we also observed a high degree of gene expression correlation for the core genome, but not the accessory genome, across all 14 patient samples (Figure 2B). This suggested the expression of core genes is conserved during human UTI, while expression of accessory genome might be more reflective of the specific conditions during each infection. Furthermore, the gene expression correlation within urine samples (Figure 2C, Figure 2D, median correlation 0.92, URINE:URINE), and within patient samples (Figure 2C, Figure 2D, median correlation 0.91, PATIENT:PATIENT) was considerably higher than the gene expression correlation between in vitro urine and patient samples (Figure 2C, Figure 2D, median correlation 0.73, URINE:PATIENT). The gene expression correlation between in vitro and patient samples remained low, even when we directly compared identical strains (i.e. HM56 cultured in vitro in urine vs. HM56 isolated from the patient) (Figure 2C, Figure 2D, median of 0.74, URINE:PATIENT:matched). This analysis suggested that UPEC adopt an infection-specific gene expression program that is distinct from UPEC undergoing exponential growth in urine in vitro. Finally, we independently confirmed this observation using principal component analysis (PCA), which revealed that patient samples form a tight cluster, distinct from in vitro cultures (Figure 2E), demonstrating the common transcriptional state of UPEC during human UTI.

Table 3.

Summary of alignment statistics (% mapped).

Sample:	Total reads	Mapped reads	% Mapped	% Mapped to CDS	% Mapped to misc_RNA	% Mapped to rRNA	% Mapped to tRNA	% Mapped to sRNA	% Mapped to tmRNA
HM01 | UR	17288419	16480326	95.3	74.91	5.51	0.01	0.26	10.2	5.49
HM01 | UTI	18496607	3717040	20.1	80.44	3.36	0	0.51	3.42	2.45
HM03 | UR	21354719	20927541	98	77.77	4.78	0	0.36	9.49	5.21
HM03 | UTI	16544044	8059076	48.7	80.18	2.45	0	0.86	2.23	1.35
HM06 | UR	23359847	22847374	97.8	78.72	3.96	0	0.33	6.3	3.23
HM06 | UTI	57993519	4709092	8.1	76.94	2.62	0	0.36	1.55	0.87
HM07 | UR	21312224	20980473	98.4	75.2	6.02	0	0.19	10.32	4.79
HM07 | UTI	70804688	2097350	3	73.71	4.14	0	0.6	2.08	0.77
HM14 | UR	21927302	21533817	98.2	76.13	5.33	0	0.15	9.97	5.16
HM14 | UTI	15944762	12968218	81.3	80.51	2.21	0	0.46	2.25	1.5
HM17 | UR	19790215	19360294	97.8	77.41	4.29	0	0.13	7.02	3.32
HM17 | UTI	23874585	1842583	7.7	74.35	4.14	0	0.73	2.73	1.6
HM43 | UR	18541484	18239826	98.4	76.54	5.03	0	0.21	9.07	4.76
HM43 | UTI	58306859	8138559	14	80.38	2.76	0	0.37	3.95	2.38
HM54 | UR	21612581	21162544	97.9	74.96	4.13	0.01	0.12	7.17	4.06
HM54 | UTI	18000843	6301998	35	77.33	3.05	0.01	0.52	1.54	0.98
HM56 | UR	17494135	17130847	97.9	77.93	4.09	0	0.09	7.14	3.56
HM56 | UTI	25408755	14935948	58.8	79.41	2.59	0	0.58	1.98	1.17
HM57 | UR	19253078	18966748	98.5	77.07	4.85	0	0.08	8.26	3.86
HM57 | UTI	105629816	926795	0.9	71.48	4.2	0	0.65	2.63	1.5
HM60 | UR	15898045	15651916	98.5	76.35	4.14	0	0.09	7.47	4.05
HM60 | UTI	76149837	764255	1	70.69	3.76	0	0.7	1.84	1.04
HM66 | UR	17184018	16736066	97.4	74.15	4.93	0	0.12	9.53	5.28
HM66 | UTI	25954183	79859	0.3	65.41	2.71	0	0.46	1.42	0.67
HM68 | UR	15841639	15562711	98.2	78.31	2.84	0	0.14	6.03	3.67
HM68 | UTI	65413931	2401089	3.7	73.11	4.8	0	0.83	4.58	2.73
HM86 | UR	15019669	14606346	97.2	76.06	4.09	0	0.16	6.99	3.54
HM86 | UTI	10667404	6413794	60.1	78.33	2.8	0	0.77	3.08	1.62

Table 4.

Summary of alignment statistics (raw counts).

Sample:	CDS	misc_RNA	rRNA	tRNA	sRNA	tmRNA
HM01 | UR	12345933	907900	1504	43435	1680592	905367
HM01 | UTI	2989889	124744	143	19133	126985	91056
HM03 | UR	16274560	999727	44	76181	1985885	1090263
HM03 | UTI	6461781	197433	24	69006	179905	109081
HM06 | UR	17985174	904287	43	76160	1439268	738927
HM06 | UTI	3623181	123428	23	17015	72873	40864
HM07 | UR	15776986	1262236	177	39363	2165537	1005391
HM07 | UTI	1546060	86761	30	12681	43708	16065
HM14 | UR	16393471	1148443	86	32625	2146180	1110769
HM14 | UTI	10441062	286490	50	59823	291189	194198
HM17 | UR	14986237	830647	48	24865	1358261	642452
HM17 | UTI	1370047	76227	15	13494	50273	29443
HM43 | UR	13960276	916836	21	37450	1653607	867656
HM43 | UTI	6541810	225003	29	30200	321597	194030
HM54 | UR	15863933	873414	1662	25326	1517844	858505
HM54 | UTI	4873058	192289	353	32932	97321	61939
HM56 | UR	13349576	701313	78	15697	1222601	609922
HM56 | UTI	11860835	386845	52	86723	295607	175048
HM57 | UR	14617905	919256	157	15069	1567276	732845
HM57 | UTI	662515	38910	13	6057	24340	13929
HM60 | UR	11949731	647306	62	13601	1169464	633959
HM60 | UTI	540215	28718	11	5361	14062	7958
HM66 | UR	12409693	825583	51	19323	1595303	884439
HM66 | UTI	52232	2161	0	366	1137	534
HM68 | UR	12187024	442312	22	22226	938831	571220
HM68 | UTI	1755457	115276	16	19970	110052	65627
HM86 | UR	11110009	597368	551	23424	1021292	517105
HM86 | UTI	5023803	179823	46	49276	197828	103919

Figure 2—figure supplement 1.

Saturation curves.

Number of mapped reads was plotted against number of expressed genes detected for each sample (in vitro samples are shown in blue; patient samples are shown in red). Vertical line shows 3 million reads cut off at which samples appear to reach saturation.

Figure 2—figure supplement 2.

Expression ranges of core genome genes.

(A) Percentage of genes in the core genome that are expressed at a given level (>1 TPM,>10 TPMs,>100 TPMs,>1000 TPMs, where TPMs are transcripts per million) is shown for patient samples that reached saturation (see Supplementary Figure 2) and corresponding in vitro samples. (B) Percentage of genes in the core genome that are expressed at a given level (>1 TPM,>10 TPMs,>100 TPMs,>1000 TPMs) is shown for patient samples that did not reach saturation and corresponding in vitro samples.

Figure 2.

Core genome expression in patients is highly correlated.

The analysis details are described in Materials and methods, and figure supplements. (A)-(B) Histogram of Pearson correlation coefficients among all samples cultured in vitro (A) or isolated from patients (B) based either on core genome or accessory genome comparisons. Accessory genome includes genes that were found in at least two but fewer than 14 of the clinical isolates. (C) Correlations among in vitro and patient samples measured by Pearson correlation coefficient of normalized gene expression plotted according to hierarchical clustering of samples. (D) Pearson correlation coefficient among all samples cultured in vitro (URINE | URINE, median = 0.92), among all samples isolated from patients (PATIENT | PATIENT, median = 0.91), between samples cultured in urine and samples isolated from patients (URINE | PATIENT, median = 0.73), and between matching urine/patient samples (ex. HM14 | URINE vs HM14 | PATIENT), (URINE | PATIENT:matched, median = 0.74). (E) Principal component analysis of normalized gene expression of 14 clinical isolates in patients and in vitro urine cultures shows distinct clustering of in vitro and patient isolates.

Number of mapped reads was plotted against number of expressed genes detected for each sample (in vitro samples are shown in blue; patient samples are shown in red). Vertical line shows 3 million reads cut off at which samples appear to reach saturation.

(A) Percentage of genes in the core genome that are expressed at a given level (>1 TPM,>10 TPMs,>100 TPMs,>1000 TPMs, where TPMs are transcripts per million) is shown for patient samples that reached saturation (see Supplementary Figure 2) and corresponding in vitro samples. (B) Percentage of genes in the core genome that are expressed at a given level (>1 TPM,>10 TPMs,>100 TPMs,>1000 TPMs) is shown for patient samples that did not reach saturation and corresponding in vitro samples.

(A) and (C) Clustering of UPEC strains cultured in filter-sterilized urine based on PCA analysis of core genome gene expression. (B) and (D) Clustering of UPEC isolated from patients based on PCA analysis of core genome gene expression. Samples in (A) and (B) are colored based on their phylogroup designation. Samples in (C) and (D) are colored based on whether the strain was isolated from a patient with recurrent UTI (Y) or without recurrent UTI (N).

We also performed PCA analysis on in vitro (Figure 2—figure supplement 3A,B) and patient samples (Figure 2—figure supplement 3C,D) separately, to ascertain whether there was any discernible effect of bacterial phylogroup (Figure 2—figure supplement 3A,C) or patients’ previous history of UTI (Figure 2—figure supplement 3B,D) on gene expression. Interestingly, B1 and B2 strains did cluster separately and a number of genes were expressed differentially in B1 and B2 backgrounds (Figure 2—source data 1, Figure 2—source data 2), suggesting that variation in gene regulatory elements between phylogroups has a small but discernible role in gene expression both in vitro and during infection. However, we found that patients’ history of UTI had no effect on bacterial gene expression.

Figure 2—figure supplement 3.

Effect of phylogenetic group on core genome expression.

(A) and (C) Clustering of UPEC strains cultured in filter-sterilized urine based on PCA analysis of core genome gene expression. (B) and (D) Clustering of UPEC isolated from patients based on PCA analysis of core genome gene expression. Samples in (A) and (B) are colored based on their phylogroup designation. Samples in (C) and (D) are colored based on whether the strain was isolated from a patient with recurrent UTI (Y) or without recurrent UTI (N).

Taken together, our data indicate diverse UPEC strains adopt a specific and conserved transcriptional program for their core genes during human infection.

UPEC show increased expression of replication and translation machinery during UTI

Differential expression analysis of the infection and in vitro transcriptomes identified 492 differentially expressed genes (log2 fold change greater than two or less than −2, adjusted p values < 0.05) (Figure 3A, Figure 3—source data 1, Figure 3—source data 2). Interestingly, pathway analysis (Table 5) and manual curation of the differentially expressed gene list (Figure 3—source data 1) revealed that expression of ribosomal subunits (r-proteins), and enzymes involved in rRNA, tRNA modification, purine and pyrimidine metabolism, and ribosome biogenesis are significantly higher in patients compared to in vitro cultures (Figure 3B). Together with previous studies (Bielecki et al., 2014; Burnham et al., 2018; Forsyth et al., 2018), these data strongly suggest that replication rates during infection are significantly higher than during mid-exponential growth in urine in vitro.

Figure 3.

Patient-associated transcriptional signature is consistent with rapid bacterial growth.

(A) The DESeq2 R package was used to compare in vitro urine cultures gene expression to that in patients. Each UPEC strain was considered an independent replicate (n = 14). Genes were considered up-regulated (down-regulated) if log2 fold change in expression was higher (lower) than 2 (vertical lines), and P value < 0.05 (horizontal line). Using these cutoffs, we identified 149 upregulated genes, and 343 downregulated genes. GO/pathway analysis showed that a large proportion of these genes belonged to one of the four functional categories (see legend). For each category, only the genes that have met the significance cut off are shown. The sugar transporters upregulated in UTI patients are shown in figure supplement. (B) Mean normalized expression for genes belonging to differentially expressed functional categories/pathways. The number of up or down-regulated genes belonging to each category is indicated next to the category name.

Heatmap shows Log2 of normalized gene expression of ptsG, fruA, fruB and gntU for each in vitro and patient sample.

Table 5.

GO modules differentially expressed in UTI patients.

Go id	Annotated	Significant	Expected	P value	Term
GO:0006518	89	24	16.63	0.03134	peptide metabolic process
GO:0016052	76	36	14.2	0.00403	carbohydrate catabolic process
GO:0044262	75	29	14.01	0.0022	cellular carbohydrate metabolic process
GO:0015980	70	20	13.08	0.02632	energy derivation by oxidation of organic compounds
GO:0043043	69	19	12.89	0.04306	peptide biosynthetic process
GO:0046395	65	25	12.14	0.00556	carboxylic acid catabolic process
GO:0006412	63	18	11.77	0.03421	translation
GO:0008643	55	30	10.28	0.02488	carbohydrate transport
GO:1903825	39	12	7.29	0.04583	organic acid transmembrane transport
GO:0008033	38	13	7.1	0.0159	tRNA processing
GO:1905039	38	12	7.1	0.03786	carboxylic acid transmembrane transport
GO:0046365	38	21	7.1	0.04177	monosaccharide catabolic process
GO:0034219	37	20	6.91	0.0005	carbohydrate transmembrane transport
GO:0042710	35	11	6.54	0.04746	biofilm formation
GO:0044010	34	11	6.35	0.03879	single-species biofilm formation
GO:0006400	34	11	6.35	0.03879	tRNA modification
GO:0072329	32	15	5.98	0.02795	monocarboxylic acid catabolic process
GO:0009401	30	11	5.6	0.01501	phosphoenolpyruvate-dependent sugar phosphotransferase system
GO:0010608	29	10	5.42	0.03121	posttranscriptional regulation of gene expression
GO:0034248	26	9	4.86	0.03925	regulation of cellular amide metabolic process
GO:0006417	26	9	4.86	0.03925	regulation of translation
GO:0015749	24	13	4.48	0.03338	monosaccharide transmembrane transport
GO:0051248	23	9	4.3	0.01728	negative regulation of protein metabolic process
GO:0044275	22	11	4.11	0.04263	cellular carbohydrate catabolic process
GO:0032269	22	8	4.11	0.03829	negative regulation of cellular protein metabolic process
GO:0015807	19	7	3.55	0.04819	L-amino acid transport
GO:0017148	18	8	3.36	0.01044	negative regulation of translation
GO:0034249	18	8	3.36	0.01044	negative regulation of cellular amide metabolic process
GO:1902475	17	7	3.18	0.02607	L-alpha-amino acid transmembrane transport
GO:0009409	14	8	2.62	0.00144	response to cold
GO:0042255	14	9	2.62	0.00021	ribosome assembly
GO:0019321	14	8	2.62	0.03705	pentose metabolic process
GO:0046835	13	6	2.43	0.02143	carbohydrate phosphorylation
GO:0006526	12	8	2.24	0.00034	arginine biosynthetic process
GO:0042542	10	5	1.87	0.02449	response to hydrogen peroxide
GO:0019323	10	7	1.87	0.02539	pentose catabolic process

We also observed infection-specific downregulation of pathways involved in amino acid biosynthesis and sugar metabolism, and a general switch from expression of sugar transporters to that of amino acid transporters (Figure 3B, Figure 3—source data 2) (with the exception of 4 sugar transporters that were expressed at higher levels in patients: ptsG, fruA, fruB, and gntU. Figure 3—figure supplement 1). Downregulation of sugar catabolism genes and upregulation of amino acid transporters suggest a metabolic switch to a more specific catabolic program as well as a scavenger lifestyle as elaborated below.

Figure 3—figure supplement 1.

Gene expression of four sugar transporters upregulated in UTI patients.

Heatmap shows Log2 of normalized gene expression of ptsG, fruA, fruB and gntU for each in vitro and patient sample.

A shift in metabolic gene expression during UTI to optimize growth potential

During our analysis, we observed that 99% (on average 2621/2653 genes) of core genome was expressed during in vitro culture, in contrast to only 94% in patient samples (2507/2653 genes). Patient samples also contained higher proportion of genes expressed at low levels when compared to in vitro samples. (Figure 2—figure supplement 2). Moreover, we noted that the majority of differentially expressed genes were downregulated in patients (343/492 differentially expressed genes). On the other hand, 30% of all upregulated genes (48/149) were ribosomal proteins. Together, these data gave us the first indication that UPEC may undergo a global gene expression reprogramming during urinary tract infection.

Bacterial growth laws postulate that bacteria dedicate a fixed amount of cellular resources to the expression of ribosomes and metabolic machinery. As a consequence, higher growth rates are achieved by allocating resources to ribosome expression at the expense of metabolic machinery production (Basan, 2018; Basan et al., 2015; Molenaar et al., 2009; Scott et al., 2010; Scott and Hwa, 2011; You et al., 2013). However, this resource reallocation between ribosomal and metabolic gene expression has not yet been measured in vivo.

First, we wanted to determine what proportion of the total transcriptome is dedicated to core genome expression. We hypothesized that during infection transcription could shift from the core genome to the accessory genome, which is enriched for virulence factors. However, we found that approximately 50% of total reads mapped to the core genome regardless of whether the bacteria were isolated from the patients or cultured in vitro (Figure 4A). Therefore, our data indicated that a fixed proportion of cellular resources were being dedicated to expression of conserved ribosomal and metabolic machinery, regardless of external environment.

Figure 4.

UPEC optimize growth potential via resource reallocation during UTI.

(A) Percentage of reads that aligned to the core genome (2653 genes) out of total mapped reads. (B) Percentage of core genome reads that mapped to r-proteins (ribosomal subunit proteins, 48 genes). (C) Percentage of core genome reads that mapped to catabolic genes (defined as genes regulated by Crp and present in the core genome (277 genes). (D) Percentage of core genome reads that mapped to amino acid biosynthesis genes (54 genes). The equivalent analysis of Subashchandrabose et al. (2014) dataset is shown in the figure supplement.

Left panel. Percentage of core genome reads that mapped to r-proteins (ribosomal subunit proteins, 48 genes) in five clinical strains from Subashchandrabose et al. study. The outlier patient sample that has only 2% of core genome mapped to r-proteins could potentially be attributed to very low depth of sequencing for that sample (HM26, see Table 6). Right panel. Percentage of core genome reads that mapped to catabolic genes. URINE: in vitro culture in filter-sterilized urine, LB: in vitro culture in LB, PATIENT: human UTI.

Table 6.

Summary of alignment statistics (% mapped) for Subashchandrabose et al. (2014).

Sample:	Total	Mapped reads	% Mapped	Mapped to CDS	Mapped to misc_RNA	Mapped to rRNA	Mapped to tRNA	Mapped to tmRNA
HM46 | UR	84195438	81447525	96.74	2.41	0.05	60.55	0.01	0.01
HM26 | UTI	20253252	1000968	4.94	16.75	0.24	21.24	0.09	0.16
HM46 | UTI	63338418	10783798	17.03	6.93	0.12	40.3	0.1	0.1
HM27 | LB	67422498	65065615	96.5	2.25	0.04	55.6	0.02	0.01
HM27 | UTI	67258748	18308171	27.22	9.25	0.13	45.49	0.08	0.2
HM26 | UR	62242978	59994538	96.39	2.31	0.08	60.58	0.01	0.01
HM65 | LB	73451346	71221338	96.96	2.53	0	51.41	0.01	0
HM69 | LB	137690758	133649727	97.07	3.49	0.05	67.26	0.01	0.01
HM69 | UTI	72509214	38506559	53.11	6.52	0.13	42.09	0.04	0.21
HM46 | LB	78018026	75590297	96.89	2.78	0.06	56.9	0.01	0.01
HM27 | UR	98185180	94683534	96.43	2.82	0.03	61	0.01	0.01
HM26 | LB	70919896	68671798	96.83	2.02	0.06	55.74	0.02	0.01
HM65 | UR	76024008	73555939	96.75	2.49	0	55.04	0.01	0
HM65 | UTI	73446576	59696718	81.28	6.19	0	40.3	0.04	0
HM69 | UR	67112750	64834311	96.61	2.45	0.04	52.92	0.01	0.01

We next looked at r-protein expression. Remarkably, we found that almost 25% of core genome reads mapped to r-proteins during infection, while this number was only 7% during exponential growth in urine (Figure 4B). These findings support the idea of extremely fast UPEC growth during UTI. Furthermore, this increase in r-protein expression correlated with a marked decrease in the proportion of core genome reads dedicated to the expression of catabolic genes (20% in vitro, 11% in patients, Figure 4C) and amino acid biosynthesis genes (5% in vitro, 1% in patients, Figure 4D). We then performed the same analysis on our previously published dataset (Subashchandrabose et al., 2014), and found a consistent trend of increased r-protein production, and decreased catabolic enzyme expression during human UTI (Figure 4—figure supplement 1, Table 6, Table 7). Thus, our data, which are consistent across multiple data sets, highlight a dramatic and conserved resource reallocation from metabolic gene expression to replication and translational gene expression during human UTI. We postulate that this resource reallocation is required to facilitate the rapid growth rate of UPEC in the host, which has been previously documented (Burnham et al., 2018; Forsyth et al., 2018).

Figure 4—figure supplement 1.

Resource reallocation analysis of Subashchandrabose et al. (2014) dataset.

Left panel. Percentage of core genome reads that mapped to r-proteins (ribosomal subunit proteins, 48 genes) in five clinical strains from Subashchandrabose et al. study. The outlier patient sample that has only 2% of core genome mapped to r-proteins could potentially be attributed to very low depth of sequencing for that sample (HM26, see Table 6). Right panel. Percentage of core genome reads that mapped to catabolic genes. URINE: in vitro culture in filter-sterilized urine, LB: in vitro culture in LB, PATIENT: human UTI.

Table 7.

Summary of alignment statistics (% mapped) for Subashchandrabose et al. (2014).

Sample	CDS	misc_RNA	rRNA	tRNA	tmRNA
HM46 | UR	1960841	36901	49312604	7302	5604
HM26 | UTI	167663	2366	212641	949	1605
HM46 | UTI	747702	12948	4345881	10289	11281
HM27 | LB	1463627	26081	36173268	11717	5088
HM27 | UTI	1693448	24245	8329004	14427	36287
HM26 | UR	1387110	48847	36345620	6532	5837
HM65 | LB	1801858	0	36612190	7263	1
HM69 | LB	4664579	71881	89896218	13828	7949
HM69 | UTI	2511733	51962	16206680	17070	81355
HM46 | LB	2099493	42356	43011663	11135	8549
HM27 | UR	2673283	31185	57757240	10152	8399
HM26 | LB	1385766	38971	38278745	11081	5724
HM65 | UR	1828039	0	40486611	5675	1
HM65 | UTI	3697360	0	24059705	24055	2
HM69 | UR	1587484	26322	34308170	4737	7686

Increase in r-protein transcripts is an infection-specific response

Doubling time during exponential growth in urine is longer than the doubling time during exponential growth in rich media, such as LB (Plank and Harvey, 1979). Thus, we wanted to determine whether the differences between the infection-specific and in vitro transcriptomes are due to longer doubling times of UPEC cultured in urine. For that purpose, one of the clinical strains, HM43, was cultured in LB, and in a new batch of filter sterilized urine. Using the growth curves shown in Figure 5A, we estimated the doubling time of HM43 during exponential growth in LB to be approximately 33 min and the doubling time in urine to be 54 min. In addition, we sequenced RNA from 3-hour-old LB cultures, 3-hour-old urine cultures and from the urine of CBA/J mice, 48 hr after transurethral inoculation with HM43 (Table 8, Table 9).

Figure 5.

Increased expression of ribosomal subunit transcripts is a host specific response.

(A) Growth curve for HM43 strain cultured in LB and filter-sterilized urine. (B) Percentage of HM43 core genome reads that mapped to ribosomal subunit proteins under different conditions (URINE: in vitro culture in filter-sterilized urine, LB: in vitro culture in LB, MOUSE: mice with UTI, PATIENT: human UTI. (C) Percentage of HM43 core genome reads that mapped to catabolic genes under different conditions.

Table 8.

Summary of alignment statistics (% mapped) for mouse UTI study.

Sample	Total reads	Mapped reads	% Mapped	Mapped to CDS	Mapped to misc_RNA	Mapped to rRNA	Mapped to tRNA	Mapped to sRNA	Mapped to tmRNA
HM43 | LB | rep1	63966646	62813946	98.2	73.01	5.49	0	0.2	11.03	6.41
HM43 | LB | rep2	37833957	37090863	98.04	71.59	5.91	0	0.2	11.63	6.69
HM43 | UR | rep1	43179946	42293006	97.95	63	8.9	0	0.06	19.96	11.94
HM43 | UR | rep2	44176952	43093840	97.55	53.64	10.94	0.01	0.03	27.8	17.9
HM43 | mouse	44314537	3690174	8.33	76.72	2.75	0	0.24	6.11	4

Table 9.

Summary of alignment statistics (% mapped) for mouse UTI study.

Sample	CDS	misc_RNA	rRNA	tRNA	sRNA	tmRNA
HM43 | LB | rep1	45862961	3449232	327	123950	6929261	4028787
HM43 | LB | rep2	26554546	2192539	204	74396	4312075	2482416
HM43 | UR | rep1	26644071	3765281	218	26488	8439668	5049595
HM43 | UR | rep2	23115456	4714597	2962	14049	11979913	7714978
HM43 | mouse	2831120	101419	55	8994	225533	147467

We then determined the proportion of r-protein transcripts in the HM43 transcriptomes isolated from urine and LB cultures. Consistent with our previous experiments, this proportion was very small in urine culture (4%). Interestingly, while the proportion of r-protein transcripts was approximately three times larger in LB cultures compared to urine, it was still significantly lower compared to what we observed during infection (Figure 5B). In contrast, the bacterial transcriptome during mouse infection exhibited r-protein expression that was similar to the human infection (Figure 5B). Additionally, the proportion of the transcriptome dedicated to catabolic gene expression was highest during urine cultures and lowest during mouse and human infections, indicating a negative correlation between the expression of r-protein and sugar catabolism genes. (Figure 5C). Overall, we show that exponential growth in rich medium alone cannot recapitulate the transcriptional signature observed during human infection. Taken together, our data suggest that the resource reallocation described in this study is an infection-specific response.

Environment-responsive regulators facilitate patient-specific gene expression program

We next sought to identify potential regulators involved in resource reallocation that facilitate the infection-specific UPEC gene expression program. To do so, we performed gene set enrichment analysis (GSEA) on E. coli co-regulated genes (regulons). This analysis allowed us to identify regulons enriched in differentially expressed genes. We identified 22 transcriptional factors whose regulon’s expression was statistically different between infection and in vitro cultures (Table 10). 18/22 regulons were expressed at higher level during in vitro culture, and eight representative regulons are shown in Figure 6. Overall, we found that these regulons accounted for 50% of differentially expressed genes that were determined to be significantly down-regulated. In contrast, only 6% of upregulated genes belonged to the four regulons that were expressed at higher levels during infection. These included genes involved in the SOS response, as well as purine synthesis (Table 10).

Table 10.

GSEA results.

Gene sets found to be enriched in differentially expressed genes. For example, Lrp, Repressor indicates gene set repressed by Lrp (data obtained from RegulonDB 9.4). Expression indicates whether regulon expression was higher in patients of during in vitro culture in urine. Regulon size: number of genes in the gene set; Matched size: number of genes found in data set; NES: normalized enrichment score; FDR: false discovery rate.

	Function	Expression (higher in)	Regulon size	Matched size	NES	FDR
Lrp	Repressor	Urine	85	27	2.29079978	0
NarL	Repressor	Urine	87	65	2.24435801	0
Lrp	Activator	Urine	38	19	2.21269565	0
MetJ	Repressor	Urine	15	14	2.12885223	0.00083422
Crp	Activator	Urine	425	277	2.12150402	0.00066738
CsgD	Activator	Urine	13	12	2.01197693	0.00250267
GadX	Activator	Urine	23	15	1.89350304	0.00929563
ModE	Activator	Urine	31	28	1.87289606	0.0108449
YdeO	Activator	Urine	18	14	1.81975146	0.02002136
Fur	Repressor	Urine	110	66	1.76658693	0.02752936
PhoP	Activator	Urine	45	33	1.7607379	0.0256334
RcsB	Activator	Urine	58	28	1.70667558	0.03781812
Hns	Repressor	Urine	144	62	1.69880665	0.03657748
GadE	Activator	Urine	70	38	1.69400478	0.03515655
RcsA	Activator	Urine	42	24	1.68615633	0.03448122
NarP	Activator	Urine	32	29	1.65675898	0.04045982
NarP	Repressor	Urine	33	26	1.6406359	0.04279074
FhlA	Activator	Urine	30	15	1.62536048	0.04514074
FliZ	Repressor	Urine	20	15	1.60948953	0.04750681
LexA	Repressor	Patients	59	43	−1.696072	0.03586007
Cra	Repressor	Patients	59	50	−1.7121855	0.04267527
PurR	Repressor	Patients	31	31	−1.752299	0.04410253
FadR	Activator	Patients	12	11	−1.9871524	0.00342544

Figure 6.

Differential regulon expression suggests role for multiple regulators in resource reallocation.

Regulon expression for 8 out of 22 regulons enriched for genes downregulated in the patients. Expression of each gene in the regulon during in vitro culture (blue) or during UTI (red) is shown along the x-axis. Histograms show proportion of genes in the regulon expressed at any given level.

In support of our previous data, the expression of catabolic genes controlled by the Crp regulator was lower in patients compared to urine cultures. In conjunction with the previously described role for Crp in resource reallocation (You et al., 2013), our in vivo findings strongly suggest that catabolite repression plays an important role in bacterial growth rate during UTI. Interestingly, other regulators identified in this analysis (NarL, ModE, MetJ, GadE, YdeO) are known sensors of environmental cues, suggesting that the infection-specific gene expression program may be driven by additional environmental signals. Taken together, we propose a model where simultaneous sensing of multiple environmental cues in the urinary tract leads to the global down-regulation of multiple metabolic regulons during infection. The cellular resources (e.g., RNA polymerase) that are freed as a result are then allocated to the transcription of genes (for example, r-proteins), which are required to maintain rapid growth rate.

Discussion

UPEC causes one of the most prevalent bacterial infections in humans; consequently, the virulence mechanisms of UPEC infection have been well-characterized. However, while we know that these virulence strategies (e.g., iron acquisition, adhesion, immune evasion) are essential for establishing infection, UPEC strains can differ dramatically in the specific factors that are utilized. Additionally, our data indicate that the expression of virulence factors can change from patient to patient, suggesting that the need for a specific factor might vary during the course of the infection.

In this study, we set out to uncover universal bacterial features during human UTIs, regardless of the stage of the infection or patient history. To do so, we performed transcriptomic analysis on bacterial RNA isolated directly from the urine of 14 patients and compared it to the gene expression of identical strains cultured to mid-exponential phase in sterile urine. Our analysis focused on the core genome as opposed to the more commonly studied accessory genome, which contains the majority of the classical virulence factors. This allowed us to identify a remarkably conserved gene expression signature shared by all 14 UPEC strains during UTI.

Although frequently overlooked, bacterial metabolism is an essential component of bacterial pathogenesis. Since the core genome is enriched for metabolic genes, we anticipated that our study would illuminate the UPEC metabolic state during human infection. Our data revealed an infection-specific increase in ribosomal protein expression in all 14 UPEC isolates, which was suggestive of bacteria undergoing rapid growth. These data strongly support the previous findings of Bielecki et al. (2014), which found a gene expression profile consistent with rapid growth in elderly patients with UTIs. Furthermore, while we did observe increased r-protein expression in exponentially growing UPEC cultured in LB, these transcripts were dramatically more abundant in the context of infection (human and mouse). Thus, the findings that UPEC maintain a conserved gene expression during UTI and grow faster in the host in comparison to in vitro conditions is consistent across multiple studies and patient cohorts (Bielecki et al., 2014), and supports recent studies that have documented very rapid UPEC growth rate measured directly in patients (Burnham et al., 2018; Forsyth et al., 2018).

Importantly, our analysis reveals how this growth rate can be achieved. We found that regardless of external environment,~50% of total gene expression is allocated to the core genome, consisting of metabolic and replication machinery, which mediate bacterial growth potential. When the infection-specific transcriptome was compared to that of UPEC cultured to mid-exponential phase in urine, we observed that elevated levels of ribosomal transcripts correlated with decreased levels of metabolic gene expression. We propose that this reallocation of resources within the core genome drives the rapid growth rate of UPEC during infection.

This resource reallocation is equivalent to what has been described as the bacterial ‘growth law’. Based on in vitro studies, the growth law proposes that increases in ribosomal gene expression occurs at the expense of a cell’s metabolic gene expression (Basan, 2018; Scott et al., 2010). Our analysis of UPEC gene expression directly from patients is consistent with this hypothesis. In addition, regulatory network analysis revealed that multiple metabolic regulons exhibit decreased transcript levels in patients suggesting an actively regulated process. In contrast, synthesis of ribosomal RNA (rRNA) coordinates the expression of ribosomal proteins by a translational feedback regulation mechanism (Jin et al., 2012; Jin and Cabrera, 2006; Nomura et al., 1984). rRNA synthesis is proposed to be regulated by the competition of RNA polymerase between transcription of rRNA operons and that of other genes, with some studies suggesting that mid-log growing cells might require almost all RNA polymerase dedicated to rRNA synthesis (Jin et al., 2012; Jin and Cabrera, 2006). Thus, decreased metabolic gene expression could allow the cell to shift its allocation of RNA polymerase towards rRNA synthesis and as a result, ribosomal protein expression. Although we cannot exclude other mechanisms, we propose that the reallocation of RNA polymerase molecules from metabolic genes to rRNA and ribosomal protein genes is a common feature adopted by diverse UPEC to promote rapid growth during UTI.

Three recent studies have attempted to characterize UPEC gene expression in patients with UTIs (Bielecki et al., 2014; Hagan et al., 2010; Subashchandrabose et al., 2014). These studies focused on the importance of virulence factor expression in specific strains and have demonstrated changes in gene expression between infection and in vitro cultures. It should be noted that all of these studies, as well as our own, were performed using bacterial RNA isolated from patient urine (that was immediately stabilized upon collection). As a result, we cannot exclude the possibility that gene expression of UPEC residing in the bladder may differ from UPEC isolated from patient urine. However, the fact remains that we and others (Bielecki et al., 2014) report that patients with different histories of UTIs all harbored a population of actively dividing bacteria in a remarkably specific metabolic state, which we have also recapitulated in a mouse model of infection in this study.

These findings raise a number of interesting questions. Firstly, how is rapid growth rate beneficial to UPEC? For example, rapid growth rate could be necessary to avoid the hosts’ innate immune response such as micturition or epithelial cell shedding. Additionally, how does this growth rate influence the tempo and mode of bacterial evolution, especially with regards to genomic integrity and the acquisition of antibiotic resistance? Finally, what are the external cues that launch the infection-specific transcriptional response? It has been noted previously that filtered urine lacks some proteins that are present in unfiltered urine (Greene et al., 2015), thus it would be interesting to see if supplementation of filtered urine with specific proteins/metabolites could recapitulate in vivo phenotype. While our study was not designed to identify infection-specific metabolites, our regulatory network analysis suggests that multiple environmental cues might reinforce the suppression of metabolic gene expression. We suggest that identifying and targeting these environmental cues is a promising approach to limit UPEC growth during UTI and gain the upper hand on this pathogen.

Materials and methods

Sample collection was previously described (Subashchandrabose et al., 2014). Briefly, a total of 86 female participants, presenting with symptoms of lower UTI at the University of Michigan Health Service Clinic in Ann Arbor, MI in 2012, were enrolled in this study. The participants were compensated with a $10 gift card to a popular retail store. Clean catch midstream urine samples from participants were immediately stabilized with two volumes of RNAprotect (Qiagen) to preserve the in vivo transcriptional profile. De-identified patient samples were assigned unique sample numbers and used in this study. Of the 86 participants, 38 were diagnosed with UPEC-associated UTIs (Subashchandrabose et al., 2014). Of these, 19 samples gave us sufficient RNA yield of satisfactory quality. Five were used for a pilot project (Subashchandrabose et al., 2014), the remaining 14 were used in this study.

Genome sequencing and assembly

The genomic DNA from clinical strains of E. coli were isolated with CTAB/phenol-chloroform based protocol. Library preparation and sequencing were performed on PacBio RS system at University of Michigan Sequencing Core. De novo assemblies were performed with canu de novo assembler (Koren et al., 2017) with all the parameters set to default mode and correction phase turned on. Finished genome assembly of reference strains (MG1655, CFT073, UTI89, EC958) were downloaded from NCBI and were converted to fastq reads using ArtificialFastqGenerator v1.0. Trimmomatic 0.36 (Bolger et al., 2014) was used for trimming adapter sequences. Variants were identified by (i) mapping filtered reads to reference genome sequence CFT073 (NC_004431) using the Burrows-Wheeler short-read aligner (bwa-0.7.17) (Li and Durbin, 2009), (ii) discarding polymerase chain reaction duplicates with Picard (picard-tools-2.5.0), and (iii) calling variants with SAMtools (samtools-1.2,) (Li, 2011) and bcftools (Li, 2011). Variants were filtered from raw results using GATK’ s (GenomeAnalysisTK-3.3–0 [Van der Auwera et al., 2013]) VariantFiltration (QUAL,>100; MQ,>50; DP >= 10 reads supporting variant; and FQ <0.025). In addition, a custom python script was used to filter out single-nucleotide variants that were <5 base pairs (bp) in proximity to indels. Positions that fell under the following regions were masked (substituted with N): (i) Phage and Repeat region of the reference genome (identified using Phaster and Nucmer; MUMmer3.23 [Kurtz et al., 2004]) (ii) Low MQ and Low FQ regions (ii) base positions that didn’t pass the hard filters (QUAL,>100; DP >= 10) were individually masked in each sample. Recombinant region identified by Gubbins 2.3.1 (Croucher et al., 2015) were filtered out and a maximum likelihood tree was constructed in RAxML 8.2.8 (Stamatakis, 2014) using a general-time reversible model of sequence evolution from the gubbins filtered alignment. Bootstrap analysis was performed with the number of bootstrap replicates determined using the bootstrap convergence test and the autoMRE convergence criteria (-N autoMRE). Bootstrap support values were overlaid on the best scoring tree identified during rapid bootstrap analysis (-f a).

Phylogroup, MLST, and serogroup typing

Phylogroups were assigned using an in-house script based on the presence and absence of primer target sequences and typing scheme (Clermont et al., 2013). MLST schemes from pubmlst (Jolley et al., 2018) were downloaded using ARIBA’s pubmlstget tool and sequence types were determined by running ARIBA (Hunt et al., 2017) against this pubmlst database. Serogroups were determined using SerotypeFinder (Joensen et al., 2015).

Bacterial culture conditions

Human urine was pooled from four age-matched healthy female volunteers. Overnight cultures of clinical isolates were washed once in human urine, then 250 μl of overnight culture was added to 25 ml of filter-sterilized human urine and cultured statically at 37C for 2 hours. Six milliliters of this culture were stabilized with RNAprotect (Qiagen) and used for RNA purification.

Antibiotic resistance profiling

As described in Subashchandrabose et al. (2014), identity and antibiotic resistance profiles of UPEC isolates were determined using a VITEK2 system (BioMerieux).

RNA isolation and sequencing

RNA isolation protocol was previously described (Subashchandrabose et al., 2014). Briefly, samples were treated with proteinase K and total RNA was isolated using Qiagen RNAeasy minikit. Turbo DNase kit (Ambion) was used to remove contaminating DNA. Bacterial content of patient samples was enriched using MICROBEnrich kit (Ambion), which depletes RNA of eukaryotic mRNA and rRNA. Library preparation and sequencing was performed by University of Michigan sequencing core. ScriptSeq Complete Kit (Bacteria) library kit was used to both deplete samples of bacterial rRNA and to construct stranded cDNA libraries from the rRNA-depleted RNA (Table 3, Table 4). While the original in vitro samples submitted for sequencing were not treated with MICROBEnrich kit, we have since performed extensive testing with two different clinical UTI strains (HM86 and HM56) to show that treatment with the kit does not affect the measured gene expression (Figure 1—figure supplement 5, Supplementary file 1). All samples were sequenced using Illumina HiSeq2500 (single end, 50 bp read length).

Figure 1—figure supplement 5.

Treatment with MICROBEnrich does not affect measures of gene expression.

Gene expression of a panel of genes was measured for HM86 (n = 3), and HM56 (n = 2) after 5 hr culture in filter-sterilized urine. After isolation RNA samples were either treated with MICROBEnrich or left untreated. Gene expression for each gene was measured by qRT-PCR. ΔCt between the gene of interest and gapA is shown.

RT-PCR validation of MICROBEnrich-treated samples

Clinical strains HM56 and HM86 were cultured overnight in LB broth at 37°C. The next morning, the culture was spun down, and the pellet washed once with PBS. Pooled filter-sterilized human urine was then inoculated with the washed bacteria at a ratio of 1:100 and incubated shaking at 37°C for five hours. Cultures were then treated with bacterial RNAprotect (Qiagen), pellets collected and stored at −80°C. The bacterial pellets were treated with both lysozyme and proteinase K, and then total RNA was extracted using the RNAeasy kit (Qiagen). Genomic DNA was removed using the Turbo DNA free kit (ThermoFisher). The extracted RNA was then halved. One half was treated using the MICROBEnrich kit (ThermoFisher), which should only remove eukaryotic mRNA and eukaryotic rRNA. The second half of the RNA remained untreated. Both the MICROBEnrich and untreated samples were reverse-transcribed into cDNA using the iScript cDNA synthesis kit (Biorad), with 1 μg RNA as template. Real-Time Quantitative Reverse Transcription PCR (qRT-PCR) was performed in a Quantstudio 3 PCR system (Applied Biosystem) in technical triplicate, using SYBR green (ThermoFisher). Samples were normalized to gapA transcript levels, by subtracting the Ct values of gapA from the Ct values of monitored genes. This value is reported as ΔCt.

Characterization of virulence factors’ gene expression

We compiled a literature search-based list of virulence factors belonging to different functional groups. Sequences for each virulence factor gene were extracted from reference UPEC genomes (either CFT073 or UTI89). Presence or absence of each virulence factor within clinical genomes was determined using BLAST (with percent identity ≥80% and percent coverage ≥90%, e-value ≤10−6). Hierarchical clustering of strains based on presence or absence of virulence factors was performed using Python’s scipy.cluster.hierarchy.linkage function with default parameters. Heatmaps of virulence factors’ gene expression in urine and in patients show normalized transcripts per million (TPMs) (same as for correlation analysis and PCA, see below).

RNAseq data processing

A custom bioinformatics pipeline was used for the analysis (Sintsova, 2019; copy archived at https://github.com/elifesciences-publications/rnaseq_analysis). Raw fastq files were processed with Trimmomatic (Bolger et al., 2014) to remove adapter sequences and analyzed with FastQC to assess sequencing quality. Mapping was done with bowtie2 aligner (Langmead and Salzberg, 2012) using default parameters. Alignment details can be found in Table 3 and Table 4. Read counts were calculated using HTseq htseq-count (union mode) (Anders et al., 2015).

Quality control

Since some of our clinical samples yielded lower numbers of bacterial reads than desired (Table 3), we performed a comprehensive quality assurance to determine if the sequencing depth of our clinical samples was sufficient for our analysis (see Saturation curves and Gene expression ranges analysis below, Figure 2—figure supplement 1, Figure 2—figure supplement 2). Overall, all patient samples except for HM66 passed quality control (see gene expression ranges analysis, Figure 2—figure supplement 2). While we elected to keep all of the strains in our subsequent analysis, this observation explains why the patient HM66 sample appears as an outlier in Figure 2.

Saturation curves

We created saturation curves for each of our sequencing files to assess whether we have sufficient sequencing depth for downstream analysis. Each sequencing file was subsampled to various degrees and number of genes detected in those subsamples (y-axis) was graphed against number of reads in the subsample (x-axis). As expected, all of the in vitro samples reached saturation (Figure 2—figure supplement 1, blue lines). Unfortunately, 6 out of our 14 samples did not reach saturation, which warranted us to investigate further (see Gene expression ranges analysis) Figure 2—figure supplement 1, red lines). Additionally, dropping the six samples that did not reach saturation from our analysis did not affect any of the results.

Core genome identification

Core genome for 14 clinical isolates and MG1655 was determined using get_homologues (Contreras-Moreira and Vinuesa, 2013). We explored multiple parameter values for our analysis and their effect on final core genome, in the end we set the cut off of 90% of sequence identity and 50% sequence coverage (similar results were obtained when using different cutoffs). The intersection of three algorithms employed by get_homologues contained 2653 gene clusters.

Gene expression ranges analysis

Due to low sequencing depth of 6 of our isolates, we were worried we would not be able to detect genes expressed at low levels in those samples. To evaluate whether we were losing information about low-level expression, we compared a number of genes in the core genome that were expressed at different levels (1000 TPMS, 100 TPMS, 10 TPMS and 1 TPM) between clinical samples that reached saturation (Figure 2—figure supplement 2A) and those that did not (Figure 2—figure supplement 2B). Only one of the clinical samples (HM66) seemed to lack genes expressed in the range of 1–10 TPMs. Thus, we conclude that all but one sample (HM66) had sufficient coverage for downstream analysis.

Pearson correlation coefficient calculation and PCA analysis

For PCA and correlation analysis, transcript per million (TPM) was calculated for each gene, TPM distribution was then normalized using inverse rank transformation. Pearson correlation and PCA was performed using python Python sklearn library. Jupyter notebooks used to generate the figures are available at https://github.com/ASintsova/HUTI-RNAseq.

Differential expression analysis

Differential expression analysis was performed using DESeq2 R package (Love et al., 2014). Genes with log2 fold change of greater than two or less than −2 and adjusted p value (Benjamini-Hochberg adjustment) of less than 0.05 were considered to be differentially expressed. DESeq2 normalized counts were used to generate Figure 3 and Figure 6. Pathway analysis was performed using R package topGO (Alexa and Rahnenfuhrer, 2018).

RNA sequencing of HM43 from the mouse model of UTI

Forty CBA/J mice were infected using the ascending model of UTI as previously described (Hagberg et al., 1983). Briefly, 40 six-week-old female mice were transurethrally inoculated with 108 CFU of UPEC isolate HM43. 48 hr post infection urine was collected from each mouse directly into bacterial RNAprotect (Qiagen). All collected urine was pooled together and pelleted, and immediately placed in the −80°C freezer. This collection was repeated every 45 minutes five more times, resulting in six collected pellets consisting of bacterial and eukaryotic cells.

For in vitro controls, UPEC strain HM43 was cultured overnight in LB. The next morning, the culture was spun down, and the pellet washed twice with PBS. LB or pooled human urine was then inoculated with the washed bacteria at a ratio of 1:100 and incubated with shaking at 37°C for 3 hr. Cultures were then treated with bacterial RNAprotect (Qiagen), pellets collected and stored at −80°C.

All the pellets were treated with both lysozyme and proteinase K, and then total RNA was extracted using RNAeasy kit (Qiagen). Genomic DNA was removed using the Turbo DNA free kit (ThermoFisher). Eukaryotic mRNA was depleted using dynabeads covalently linked with oligo dT (ThermoFisher). The in vitro samples underwent the same treatment with dynabeads to reduce any potential biases this procedure might introduce to the downstream sequencing. The supernatant was collected from this treatment, and the RNA was concentrated and re-purified using RNA Clean and Concentrator kit (Zymo). Library preparation and sequencing was performed by University of Michigan sequencing core. The ScriptSeq Complete Gold Kit (Epidemiology) library kit was used to both deplete samples of bacterial and eukaryotic rRNA and to construct stranded cDNA libraries from the rRNA-depleted RNA. These were sequenced using Illumina HiSeq2500 (single end, 50 bp read length). RNAseq analysis was performed as described above, alignment statistics are shown in Table 8 and Table 9.

Analysis of RNAseq data from Subashchandrabose et al. (2014). Sample collection and RNA isolation is described in Subashchandrabose et al. (2014). Briefly, RNA samples were treated with proteinase K and total RNA was isolated using Qiagen RNAeasy minikit. Turbo DNase kit (Ambion) was used to remove contaminating DNA. Bacterial content of patient samples was enriched using MICROBenrich kit (Ambion). The depleted RNA was used to generate sequencing libraries using the Ovation Prokaryotic RNA-Seq system (NuGen) and the Encore next-generation sequencing library system (NuGen). The libraries were sequenced using an Illumina HiSeq2000 (paired-end, 100 bp) by the Genome Resource Center at the Institute for Genome Sciences, University of Maryland, Baltimore, MD. RNAseq analysis was performed as described above, alignment statistics are shown in Table 6 and Table 7.

Estimation of HM43 doubling time

For both LB and urine OD curves were performed using Bioscreen-C Automated Growth Curve Analysis System (Growth Curves USA) eight separate times. For each time point, the mean values of the eight replicates were used for doubling time estimation. The equation bellow was used to estimate doubling time during logarithmic growth in LB or urine, where DT is doubling time, C2 is final OD, C1 is initial OD, and  is time elapsed between when C2 and C1 were taken.

DT was calculated for every two measurements taken between 30 and 180 min and mean of these values is reported.

Regulon analysis

Regulon gene sets were extracted from RegulonDB 9.4 (Gama-Castro et al., 2016) using custom Python scripts (available https://github.com/ASintsova/HUTI-RNAseq). Gene set enrichment analysis was performed using Python GSEAPY library.

Data access

Jupyter notebooks as well as all the data used to generate the figures in this paper are available on github: https://github.com/ASintsova/HUTI-RNAseq.

In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.

[Editors’ note: a previous version of this study was rejected after peer review, but the authors submitted for reconsideration. The first decision letter after peer review is shown below.]

Thank you for submitting your work entitled "Genetically diverse Escherichia coli adopt a common transcriptional program in patients with urinary tract infections" for consideration by eLife. Your article has been reviewed by three peer reviewers, and the evaluation has been overseen by a Reviewing Editor and a Senior Editor. The following individual involved in review of your submission has agreed to reveal their identity: Alexander Westermann (Reviewer #2).

Our decision has been reached after consultation between the reviewers. Based on these discussions and the individual reviews below, we regret to inform you that your work will not be considered further for publication in eLife. All reviewers agreed on the importance of your work on characterising the transcriptional program of strains directly from patients. However, they raised serious technical questions with respect to the different processing of the RNA samples from patients, mice and in vitro, and potential confounding factors that cannot be eliminated without extensive new experiments. If you wished to address all the reviewers' concerns, we would be willing to assess the suitability for publication of an extensively revised version of this manuscript.

Reviewer #1:

This manuscript continues a large body of work by Mobley and co-workers to define a UPEC virulence genotype during human UTI. The authors use RNAseq to characterise a transcriptional program shared by a genetically diverse group of 14 UPEC strains harvested directly from the urine of infected patients. This revealed a signature defining upregulation of genes involved in translation, and demonstrating that UPEC grow rapidly during human UTI.

A major strength of this work is the transcriptome analysis of UPEC growth during human UTI, and comparison of this to growth in vitro and growth in the mouse model of UTI. The conserved transcriptional signature of the core genome during human UTI led to the important discovery that reprogramming occurs, and results in the allocation of cellular resources to support rapid growth during human UTI. It was comforting to see a conserved profile of transcription in humans and mice, an important finding that will be of great value to the field. In a broader context, this work provides new insight into UPEC adaptation to the human urinary tract. The transcriptional program that defines rapid growth during human UTI provides a framework to design novel therapeutics that block this phenotype in UPEC, an urgent need in the context of rapidly increasing antibiotic resistance.

Overall, I have no substantive concerns regarding the data or major conclusions.

Reviewer #2:

In the present study, Sintsova et al. profile expression of the core genome of 14 UPEC isolates right upon their collection from UTI patients and compare it with the corresponding bacterial transcriptome patterns when grown in vitro. In doing so, the authors identify a conserved expression pattern shared among all isolates, which is associated with an increased expression of mRNAs encoding ribosomal proteins and a reduced expression of metabolic genes in the in vivo isolates compared to the in vitro cultures. Based on this finding, the authors conclude that in vivo UPEC reallocates its resources to increase proliferation at the expense of metabolic activity, and consequently, that bacterial growth may be enhanced in the patient's bladder as compared to in vitro conditions. This would imply that one or several elusive host factors (present in the bladder but absent from the sterile-filtered urine) would enhance UPEC growth.

My major concern is that the study lacks enough data to support this hypothesis. The growth rate is deduced from the number of sequencing reads that map to genes for ribosomal proteins. This measure happens to fit with the growth behavior in vitro, when the authors compare UPEC replication in LB and sterile urine. However, this does not necessarily mean that the same applies in vivo. Given that this is their major finding, the authors may want to further support this speculation by in vivo data, e.g. by determining CFU counts in their mouse model over time of infection. Also, what would speak against their hypothesis is that genes for other cellular functions required for growth (such as DNA replication, cell division, etc.) appear not to be differentially expressed in vivo vs. in vitro.

Also with respect to the generation and analysis of RNAseq data, I have some comments:

– It remains unclear if the authors sequenced bacterial total RNA or rRNA-depleted RNA. The MICROBEnrich kit used, depletes polyadenylated transcripts (eukaryotic mRNAs and certain lincRNAs) as well as eukaryotic rRNAs, but does not efficiently deplete bacterial rRNA. Also the Illumina ScriptSeq v2 kit per se, does not deplete ribosomal transcripts. Therefore, it appears unclear why the authors state that "… rRNA-depleted stranded cDNA libraries…" were constructed. By the way, if indeed not actively depleted, reads mapping to bacterial rRNAs should also be increased in the in vivo samples (as are reads mapping to ribosomal proteins). In general, plots or tables that inform about RNA class distributions (% reads mapping to mRNAs, rRNAs, tRNAs, etc.) in the individual samples would be helpful.

– Why didn't the authors include published datasets in their analysis? This would seem particularly obvious for the data derived from their own previous (pilot) study (Subashchandrabose et al., 2014) that was based on 5 samples taken together with the 14 samples analyzed here.

Reviewer #3:

The manuscript "Genetically diverse uropathogenic Escherichia coli adopt a common transcriptional program in patients with urinary tract infections", by Sintsova et al., presents an RNAseq-driven analysis of UPEC gene expression from 14 UTI patients. The main experiment is to compare the expression of the bacteria directly isolated from the patients to the expression of the bacteria after growth in filter-sterilized, pooled human urine in vitro. The primary result is that bacteria isolated from the urine of patients have high expression of genes encoding proteins involved in DNA and protein synthesis: ribosomal proteins, rRNA and tRNA modification proteins, purine and pyrimidine metabolism.

The primary result was examined with a few analyses on the patient vs. in vitro urine expression. The analysis was split early between virulence genes/accessory genes and "core" genes that are present in all the strains. The analysis focused relatively quickly on the core genes, which included all the DNA and protein synthesis genes noted above. Further analysis of the core genes resulted in a few other general features that differentiated patient (infection) from in vitro growth: downregulation of amino acid biosynthesis, downregulation of sugar metabolism, downregulation of most sugar transporters, and upregulation of amino acid transporters.

An important set of validation experiments was then done with one strain, HM43, which was grown in LB (considered a "rich" lab media), a new batch of filter-sterilized urine, and multiple urine samples from mice that had been infected in their bladders with HM43. Using% of reads mapping to genes encoding ribosomal proteins and to genes encoding catabolic enzymes showed that the in vivo mouse infection samples again had a high proportion of ribosomal reads and low catabolic reads, more similar to the human patient expression data and not matched by the LB-grown bacteria. This last comparison was the main test the effect of a faster growth rate per se.

The authors then conclude that there is an infection-specific transcriptional program which is dedicated to high growth rate in urine. They further noted that the downstream regulated genes for 22 transcriptional factors were differentially expressed between patient infections and in vitro urine growth, and speculate that some environmental cues may be sensed and thereby lead to the observed high expression of DNA and protein synthesis genes.

I have one primary technical question about the experimental design and a question about the context within the UTI field. First, the authors have had substantial experience with doing RNAseq from patient urines, and take care to attempt to minimize the time between urine sample collection and RNAProtect addition (to be applauded). Also, all samples according to the Materials and methods are also stored in RNAProtect. One remaining issue is that it states that the "bacterial content of patient samples was enriched using MICROBEnrich kit". This raises a potential confounding variable that seems reasonable for explaining a large scale consistent different between patient samples and in vitro urine samples. For the HM43 mouse experiment, it is stated that "eukaryotic mRNA was depleted using dynabeads covalently linked with oligodT". This would seem to leave the eukaryotic ribosomal RNA still in the sample, but sequencing and mapping statistics for this experiment (similar to Table 1 for the human samples) are not included to check on this. Therefore, I am wondering whether the in vivo mouse samples were also treated differently than the in vitro LB and urine samples. Regardless, the authors should clarify the methods particularly for whether the MICROBEnrich was used only on the patient samples and explicitly not on the in vitro urine samples, and also they should similarly provide a bit more detail on the processing of the mouse samples.

In addition, the primary result from this paper has largely been described before. Bielecki et al. (2014) performed RNAseq on 21 strains from human UTI patients, 4 of which were subsequently grown in LB for to get an in vitro RNAseq data set; also of note these authors used MICROBEnrich for host RNA removal on the human patient samples, and they did rRNA depletion with a MICROBExpress kit for the in vitro samples. I suggest that the authors more explicitly acknowledge this previous work and the general observations that have already been made, which will help them to either better differentiate their current study or strengthen the overall result shared by the two papers by providing additional validating data.

[Editors’ note: what now follows is the decision letter after the authors submitted for further consideration.]

Thank you for submitting your article "Genetically diverse uropathogenic Escherichia coli adopt a common transcriptional program in patients with UTIs" for consideration by eLife. Your article has been reviewed by three peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Neil Ferguson as the Senior Editor. The following individual involved in review of your submission has agreed to reveal their identity: Alexander Westermann (Reviewer #2).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

The work presented here is a revised version of a previous submission. The original paper was deemed very interesting by reviewers, who however had raised several key issues, notably potential confounding of results due to differences in in vitro vs. in vivo RNA isolation/processing in the lab. The reviewers now agree that you have provided substantial new data to address these concerns and would be willing to consider this manuscript positively for publication. However, there are a list of points that should be addressed before this could be considered.

Essential revisions:

1) Could you explain how you selected the extra dataset provided in Figure 1—figure supplement 5 that suggests that MICROBEnrich has little effect on bacterial gene expression? Importantly, were these some of the genes that showed differential expression between the in vivo and in vitro samples? It is essential to include, if not already tested, at least some of the r-protein-encoding mRNAs in this analysis given their importance in the study.

2) Table 3: Only a small percentage of the reads from the in vivo samples map to the UPEC genome. Could you explain what the remainder of the reads might be derived from? Would they map to the host genome (thus indicating that bacterial enrichment is rather low-efficient), or is there any evidence that some of those non-UPEC reads might be derived from other causative agents of UTI?

[Editors’ note: the author responses to the first round of peer review follow.]

Reviewer #2:

[…] My major concern is that the study lacks enough data to support this hypothesis. The growth rate is deduced from the number of sequencing reads that map to genes for ribosomal proteins. This measure happens to fit with the growth behavior in vitro, when the authors compare UPEC replication in LB and sterile urine. However, this does not necessarily mean that the same applies in vivo. Given that this is their major finding, the authors may want to further support this speculation by in vivo data, e.g. by determining CFU counts in their mouse model over time of infection. Also, what would speak against their hypothesis is that genes for other cellular functions required for growth (such as DNA replication, cell division, etc.) appear not to be differentially expressed in vivo vs. in vitro.

We appreciate the reviewers concerns and apologize for lack of clarity in our discussion of data on growth rate:

1) As mentioned by the reviewer, levels of ribosomes/rRNA are known to be closely correlated with growth rate under many different experimental conditions (Basan, 2018; Basan et al., 2015; Molenaar et al., 2009; Scott et al., 2010; Scott and Hwa, 2011; You et al., 2013), and there is no reason to believe that this would not also hold true in an in vivosetting. Additionally, while, in the past, we have measured CFUs during infection, bacterial load cannot be used as a proxy for growth rate, as it will be affected by multiple factors besides bacterial growth rate (such as bacterial clearance by the immune response to the infection).

2) We apologize that this was not clear from our text, but in fact we do show that other genes required for growth are upregulated during infection (Figure 3B, subsection “UPEC show increased expression of replication and translation machinery during UTI”, first paragraph). In addition to genes mentioned in Figure 3, there are other genes required for growth that are listed in Figure 3—source data 1, such as fis, dnaG (DNA replication), dbpA (ribosome biogenesis), mrcA, and mrdA (peptidoglycan biosynthesis). Additionally, while there are sure to be other genes/proteins that are required for growth that were not significantly upregulated during infection, they might be regulated at a translational/post-translational level, and thus would not be changed in their expression.

3) Our conclusions about rapid growth rate are not only based on the gene expression profile observed in this study, but are also supported by previous studies that attempt to directly measure UPEC growth rate in patients with UTIs (Burnham et al., 2018; Forsyth et al., 2018). We have tried to make this clear throughout the manuscript (specifically in the last paragraph of the Introduction, the first paragraph of the subsection “UPEC show increased expression of replication and translation machinery during UTI” and in the third paragraph of the Discussion). In this study we show how this rapid growth rate can result from repartitioning of the bacterial transcriptome that is observed both during mouse and human UTI.

Also with respect to the generation and analysis of RNAseq data, I have some comments:

– It remains unclear if the authors sequenced bacterial total RNA or rRNA-depleted RNA. The MICROBEnrich kit used, depletes polyadenylated transcripts (eukaryotic mRNAs and certain lincRNAs) as well as eukaryotic rRNAs, but does not efficiently deplete bacterial rRNA. Also the Illumina ScriptSeq v2 kit per se, does not deplete ribosomal transcripts. Therefore, it appears unclear why the authors state that "… rRNA-depleted stranded cDNA libraries…" were constructed. By the way, if indeed not actively depleted, reads mapping to bacterial rRNAs should also be increased in the in vivo samples (as are reads mapping to ribosomal proteins). In general, plots or tables that inform about RNA class distributions (% reads mapping to mRNAs, rRNAs, tRNAs, etc.) in the individual samples would be helpful.

We apologize for the lack of clarity and omission of RNA class distribution data. We have revised our Materials and methods section (subsection “RNA isolation and sequencing”) and added Tables 3 and 4 to address this point. These tables show clear depletion of bacterial rRNA from our samples. Specifically, both eukaryotic and prokaryotic rRNA was depleted from the samples prior to sequencing. Bacterial content of patient samples was enriched using MICROBEnrich kit (Ambion), which depletes the sample of eukaryotic mRNA and rRNA. Library preparation and sequencing was performed by University of Michigan sequencing core. ScriptSeq Complete Kit (Bacteria) library kit (https://www.illumina.com/products/scriptseq-bacteria.html) was used to both deplete samples of bacterial rRNA and to construct stranded cDNA libraries from the rRNA-depleted RNA.

– Why didn't the authors include published datasets in their analysis? This would seem particularly obvious for the data derived from their own previous (pilot) study (Subashchandrabose et al., 2014) that was based on 5 samples taken together with the 14 samples analyzed here.

The samples from our pilot study were not analyzed alongside this study because the library preparation, sequencing and facility used in the pilot study were different. However, we appreciate this comment, which has also been brought up by reviewer #3. We have now analyzed data from Subashchandrabose (2014) and show that it is consistent with our current study, i.e., increase intranscripts dedicated to r-protein production in the patient samples compared to LB or urine. These data have now been added as Figure 4—figure supplement 1 and are discussed in the last paragraph of the subsection “A shift in metabolic gene expression during UTI to optimize growth potential”, and in the subsection “Analysis of RNAseq data from Subashchandrabose et al.” One of the patient samples (HM26) had very few reads that mapped to the bacterial genome (Table 6, 7), which potentially explains the fact that this was the only patient sample that contained very few r-protein reads.

Reviewer #3:

[…] I have one primary technical question about the experimental design and a question about the context within the UTI field. First, the authors have had substantial experience with doing RNAseq from patient urines, and take care to attempt to minimize the time between urine sample collection and RNAProtect addition (to be applauded). Also, all samples according to the Materials and methods are also stored in RNAProtect. One remaining issue is that it states that the "bacterial content of patient samples was enriched using MICROBEnrich kit". This raises a potential confounding variable that seems reasonable for explaining a large scale consistent different between patient samples and in vitro urine samples. For the HM43 mouse experiment, it is stated that "eukaryotic mRNA was depleted using dynabeads covalently linked with oligodT". This would seem to leave the eukaryotic ribosomal RNA still in the sample, but sequencing and mapping statistics for this experiment (similar to Table 1 for the human samples) are not included to check on this. Therefore, I am wondering whether the in vivo mouse samples were also treated differently than the in vitro LB and urine samples. Regardless, the authors should clarify the methods particularly for whether the MICROBEnrich was used only on the patient samples and explicitly not on the in vitro urine samples, and also they should similarly provide a bit more detail on the processing of the mouse samples.

1) We sincerely apologize for the lack of clarity in our Materials and methods, as we did not state clearly whether eukaryotic/prokaryotic rRNA was depleted prior to RNA sequencing. We have expanded the Materials and methods section to clearly state that indeed both eukaryotic and prokaryotic rRNA was depleted from all of our samples sequenced in this study (subsection “RNA isolation and sequencing”), and provide additional tables showing numbers of reads that mapped to different types of RNA for each sample, for both clinical, in vitro and mouse samples (Tables 3, 4, 8, and 9).

2) Additionally, we also failed to clearly explain that for the mouse study described in Figure 5 all samples (in vitroLB and urine cultures, as well as RNA isolated from mice with UTIs) were subjected to the same protocol (i.e., treatment with dynabeads covalently linked to oligo dT) to avoid any potential biases that could result from differential RNA processing. This clarification has now been added in the last paragraph of the subsection “RNA sequencing of HM43 from mouse model of UTI”.

3) The reviewer brought up a very important point suggesting that treatment of patient samples with MICROBEnrich could result in differential gene expression program we observed in the study. We want to point out that if treatment with MICROBEnrich affected the measurement of gene expression or RNA pool composition, we would expect to see large-scale global difference in gene expression, when in fact less than 500 genes were differentially expressed between the two conditions and were enriched in specific biological processes. However, we have taken this concern very seriously and provide five separate lines of evidence that the gene expression pattern observed in patient population is not due to an RNA processing step. These are summarized above, and are further elucidated below:

- We have performed an extensive RT-PCR study to show that treatment with MICROBEnrich does not change the levels of gene expression in multiple UPEC strains cultured in urine. Specifically, we isolated RNA from urine cultures of HM86 and HM56, split each of the samples in half, where one half went through MICROBEnrich treatment, and the other was left untreated. We then measured gene expression for a panel of 10 different genes for both treated and untreated samples by RT-PCR and found no differences in gene expression between any of them. This was done in three biological replicates for HM86, and two biological replicates for HM56. This provides strong evidence that treatment with MICROBEnrich alone does not affect measurement of gene expression. These experiments are now explained in the Materials and methods subsection “RT-PCR validation of MICROBEnrich-treated samples” and presented in Figure 1—figure supplement 5, Table 10, and Supplementary file 1.

- As discussed above, for the mouse UTI experiments, in vitroand in vivo RNA samples were treated exactly the same (yet differently from the original urine and patient samples). Nevertheless, the gene expression during mouse UTI highly resembles that of human UTI. This excludes the possibility of our results being attributable to MICROBEnrich.

- In the revised Figure 1, we take a closer look at expression of genes that we expect to be similarly highly expressed in urine and patients (i.e., iron acquisition genes). As expected, we see a similar pattern of expression for a number of different virulence factors between in vitroand patient samples, which argues against a MICROBEnrich treatment having an effect on measured gene expression. This data is now presented in Figure 1B and Figure 1—figure supplement 3 and is described in the Results subsection “Virulence factor expression is not specific to infection”.

- One of the main conclusions of the paper (i.e., conserved gene expression program during UTI) does not rely on the comparison between in vitroand patient samples, and therefore the difference in treatment between these two sample groups would have minimal impact on our stated results. In fact, we first observed the conserved gene expression program when we compared gene expression of core genome and accessory genome between patients and observed that the core gene expression is more conserved. We apologize for omitting these data from the original manuscript, it is now included in Figure 2A and 2B and is discussed in the first paragraph of the subsection “The UPEC core genome exhibits a common gene expression program during clinical Infection”.

- Finally, we believe our original manuscript did not stress enough the importance of previous work on UPEC growth rate during infection. Specifically, Bielecki et al.(2014) have also observed a gene expression pattern consistent with fast growth in patients with UTIs. In addition, two other studies attempt to directly measure UPEC growth rate in patients with UTI and find it to be consistently fast [(Burnham et al., 2018; Forsyth et al., 2018), the latter from our lab]. The fact that multiple studies across different patient cohorts came to similar conclusions argues against our study being affected by differential MICROBEnrich treatment. We have revised the manuscript to highlight the previous work done by us and others throughout the Results and Discussion. Specifically, the following was modified: Introduction, last paragraph; subsection “UPEC show increased expression of replication and translation machinery during UTI”, first paragraph; Discussion, third and sixth paragraphs.

In addition, the primary result from this paper has largely been described before. Bielecki et al. (2014) performed RNAseq on 21 strains from human UTI patients, 4 of which were subsequently grown in LB for to get an in vitro RNAseq data set; also of note these authors used MICROBEnrich for host RNA removal on the human patient samples, and they did rRNA depletion with a MICROBExpress kit for the in vitro samples. I suggest that the authors more explicitly acknowledge this previous work and the general observations that have already been made, which will help them to either better differentiate their current study or strengthen the overall result shared by the two papers by providing additional validating data.

As discussed above, we are aware of the importance of the study described here, and it was not our intention to discount it in any way. In fact, we think that the fact that the two studies are in close agreement, despite studying different patient cohorts, is remarkable and significantly strengthens the conclusions of both studies. We have revised the text of the manuscript to more explicitly acknowledge this work, specifically the following was modified: Introduction, last paragraph; subsection “UPEC show increased expression of replication and translation machinery during UTI”, first paragraph; Discussion, third and sixth paragraphs. We also want to emphasize that the purpose of our study was also to build on these findings and provide a mechanistic insight into how this growth rate could potentially be achieved. Specifically, we show how differential transcriptome partitioning can facilitate the rapid growth. Moreover, we recapitulate our observations in the mouse model of UTI and characterize E. coli regulatory network that is behind the transcriptome partitioning observed during human UTI.

[Editors' note: the author responses to the re-review follow.]

Essential revisions:

1) Could you explain how you selected the extra dataset provided in Figure 1—figure supplement 5 that suggests that MICROBEnrich has little effect on bacterial gene expression? Importantly, were these some of the genes that showed differential expression between the in vivo and in vitro samples? It is essential to include, if not already tested, at least some of the r-protein-encoding mRNAs in this analysis given their importance in the study.

For the original panel, we selected genes that have previously been shown to be important for UPEC infections. However, they were not differentially expressed between in vivoand in vitroconditions. We have now extended this panel by including data for five more genes, 2 of which (nanM and malK) were downregulated under in vivo conditions, and 3 of which (ribosomal protein genes rpoA, rplA, rpsA) were upregulated in vivo. As is shown in Supplementary file 1, MICROBEnrich treatment did not affect measurements of gene expression for any of the genes in the panel.

2) Table 3: Only a small percentage of the reads from the in vivo samples map to the UPEC genome. Could you explain what the remainder of the reads might be derived from? Would they map to the host genome (thus indicating that bacterial enrichment is rather low-efficient), or is there any evidence that some of those non-UPEC reads might be derived from other causative agents of UTI?

The majority of the reads from those samples do indeed map to the host genome. Moreover, none of the patients in this study had polymicrobial infections. While the enrichment was not perfect, in our hands, it does significantly increase the amount of bacterial RNA in our samples.

Key resources table

Reagent type (species) or resource	Designation	Source or reference	Identifiers	Additional information
Strain, strain background (Escherichia coli)	Uropathogenic Escherichia coli HM01	This study		Strain isolation described in Study Design section below
Strain, strain background (Escherichia coli)	Uropathogenic Escherichia coli HM03	This study		Strain isolation described in Study Design section below
Strain, strain background (Escherichia coli)	Uropathogenic Escherichia coli HM06	This study		Strain isolation described in Study Design section below
Strain, strain background (Escherichia coli)	Uropathogenic Escherichia coli HM07	This study		Strain isolation described in Study Design section below
Strain, strain background (Escherichia coli)	Uropathogenic Escherichia coli HM14	This study		Strain isolation described in Study Design section below
Strain, strain background (Escherichia coli)	Uropathogenic Escherichia coli HM17	This study		Strain isolation described in Study Design section below
Strain, strain background (Escherichia coli)	Uropathogenic Escherichia coli HM43	This study		Strain isolation described in Study Design section below
Strain, strain background (Escherichia coli)	Uropathogenic Escherichia coli HM54	This study		Strain isolation described in Study Design section below
Strain, strain background (Escherichia coli)	Uropathogenic Escherichia coli HM56	This study		Strain isolation described in Study Design section below
Strain, strain background (Escherichia coli)	Uropathogenic Escherichia coli HM57	This study		Strain isolation described in Study Design section below
Strain, strain background (Escherichia coli)	Uropathogenic Escherichia coli HM60	This study		Strain isolation described in Study Design section below
Strain, strain background (Escherichia coli)	Uropathogenic Escherichia coli HM66	This study		Strain isolation described in Study Design section below
Strain, strain background (Escherichia coli)	Uropathogenic Escherichia coli HM68	This study		Strain isolation described in Study Design section below
Strain, strain background (Escherichia coli)	Uropathogenic Escherichia coli HM86	This study		Strain isolation described in Study Design section below
Strain, strain background (Escherichia coli)	Uropathogenic Escherichia coli HM26	(Subashchandrabose et al., 2014)
Strain, strain background (Escherichia coli)	Uropathogenic Escherichia coli HM27	(Subashchandrabose et al., 2014)
Strain, strain background (Escherichia coli)	Uropathogenic Escherichia coli HM46	(Subashchandrabose et al., 2014)
Strain, strain background (Escherichia coli)	Uropathogenic Escherichia coli HM65	(Subashchandrabose et al., 2014)
Strain, strain background (Escherichia coli)	Uropathogenic Escherichia coli HM69	(Subashchandrabose et al., 2014)
Strain, strain background (Mus musculus)	CBA/J
commercial assay or kit	MICROBEnrich Kit	Thermo Fisher	AM1901
commercial assay or kit	RNeasy kit	Qiagen	74104
commercial assay or kit	Turbo DNase kit	Thermo Fisher	AM2238
commercial assay or kit	iScript cDNA synthesis kit	Bio Rad	1708890
commercial assay or kit	ScriptSeq Complete Gold Kit (Epidemiology)	Illumina	Discontinued
commercial assay or kit	ScriptSeq Complete Kit (Bacteria)	Illumina	Discontinued
commercial assay or kit	PowerUP SYBR Green Master Mix	Bio Rad	A25779
commercial assay or kit	Dynabeads mRNA DIRECT Purification kit	Thermo Fisher	61011
chemical compound, drug	RNAprotect	Qiagen	76526
software, algorithm	Trimmomatic	(Bolger et al., 2014)	0.36
software, algorithm	Bowtie2	(Langmead and Salzberg, 2012)	2.3.4
software, algorithm	samtools	(Li, 2011)	1.5
software, algorithm	HTseq	(Anders et al., 2015)	0.9.1
software, algorithm	Get_homologues	(Contreras-Moreira and Vinuesa, 2013)	20170807
software, algorithm	DESeq2	(Love et al., 2014)	1.22.2