Plasmodium falciparum serology: A comparison of two protein production methods for analysis of antibody responses by protein microarray.

The evaluation of protein antigens as putative serologic biomarkers of infection has increasingly shifted to high-throughput, multiplex approaches such as the protein microarray. In vitro transcription/translation (IVTT) systems-a similarly high-throughput protein expression method-are already widely utilised in the production of protein microarrays, though purified recombinant proteins derived from more traditional whole cell based expression systems also play an important role in biomarker characterisation. Here we have performed a side-by-side comparison of antigen-matched protein targets from an IVTT and purified recombinant system, on the same protein microarray. The magnitude and range of antibody responses to purified recombinants was found to be greater than that of IVTT proteins, and responses between targets from different expression systems did not clearly correlate. However, responses between amino acid sequence-matched targets from each expression system were more closely correlated. Despite the lack of a clear correlation between antigen-matched targets produced in each expression system, our data indicate that protein microarrays produced using either method can be used confidently, in a context dependent manner, though care should be taken when comparing data derived from contrasting approaches.

Introduction

To date, the majority of malaria serologic studies have focussed on antibody responses to a small number of well-characterised, highly immunogenic Plasmodium falciparum antigens that have proven to be reliable markers of exposure to infection [1-8]. However, P. falciparum expresses more than 5000 proteins, each a potential antibody target [9, 10]. Advances in technology have led to the development of new assay platforms that allow proteome scale investigation of antibody responses, such as the protein microarray [11, 12]—boasting significantly greater experimental throughput than more classical monoplex methods (e.g. ELISA) [13, 14]. The ability to simultaneously interrogate large numbers of putative targets, using low volumes of sample, significantly increases the rate at which an individual’s antibody responses to antigens can be characterised. As such, protein microarray based approaches to biomarker identification and humoral response profiling in malaria, and other infectious diseases, have been increasingly adopted [15-24].

One widely utilised form of the protein microarray is based on an in vitro transcription/ translation (IVTT) system [25]–where protein products are produced through a PCR, in vivo recombination cloning and an in vitro expression pipeline, before being printed onto arrays [15]. In principle, whole organism proteome microarrays can be fabricated simply and quickly, enabling analysis of all potential protein driven immune responses to a pathogen. Cell-free synthesis (CFS) is a technique first established over 50 years ago as a means to dissect the molecular mechanisms around protein expression. More recently, the technique has been used as a high throughput expression platform to explore a number of diverse biological processes [26, 27]. At its simplest, the approach utilises the crude extract containing the transcription and translation machinery from the cell, performing the process of protein expression without the constraints of the cell. This allows a wide variety of proteins to be expressed including those that would be deemed toxic if expression was attempted within the confines of the cell membrane [28]. CFS systems based on Escherichia coli (E.coli) are among the most widely used of the IVTT systems [27] and have helped to transform the narrative around a number of areas including biomarker discovery for infectious diseases [15, 29, 30]. Despite the widespread uptake of the approach there remain some issues around the technique. This includes significant heterogeneity of expression, leading some research groups to describe the mechanisms of the process as a “black box”. Therefore, the inherent heterogeneity between products is not assessed for every target making it difficult to normalise for reactivity between protein spots, which represent an impure mix of E. coli and target protein. In addition to the E. coli cell-free expression platform, other approaches have been employed in the characterisation of protein targets for immunological assessment. The wheat germ cell-free expression system in particular has also proven to be an important platform in the advancement of biomarker discovery and malaria vaccine research [31-34]. This is not the focus of the current study.

In contrast to the IVTT array methodology, the printing of purified proteins is cheaper and typically more quantifiable. Uniform amounts of product can therefore be incorporated into arrays, increasing confidence when comparing quantitative antibody responses between antigenic targets [35] and assessing relative immunogenicity. The process can be modified to support the scale up of recombinant proteins, and furthermore, affinity purification of protein targets reduces the risk of undesired background reactivity due to expression system components, and in part truncated proteins. However, the time required to produce panels of purified proteins is far in excess of the IVTT system, particularly for large numbers of targets, unless supported by an automated production platform [36-38]. For both the IVTT and purified protein E. coli systems, although the production of complex conformational proteins is possible it can sometimes be a challenge [39, 40]. These challenges are in part due to the expression of proteins foreign to the bacteria, the speed at which bacteria express proteins, only partially mitigated with a reduction in expression temperature; and the lack of essential molecular chaperones to aid correct folding/refolding of proteins [41-43].

Here we present a comparison between IVTT based and purified proteins on a single microarray. For clarity proteins produced using the IVTT system will simply be referred to as IVTT proteins, and those produced by conventional E.coli expression will be referred to as purified proteins. Matched malarial protein targets from each methodology were assessed for comparative reactivity in serum from Ugandan participant samples (n = 899) [44] to determine the suitability of each approach in the context of high-throughput profiling of serological responses to protein antigens.

Material and methods

Ethics statement

All serum samples were collected after written informed consent from the participant or their parent/guardian. The protocol for sample collection was reviewed and approved by the Makerere University School of Medicine Research and Ethics Committee (#2011–149 and #2011–167), the London School of Hygiene and Tropical Medicine Ethics Committee (#5943 and #5944), the Durham University School of Biological and Biomedical Sciences Ethics Committee, the University of California, San Francisco, Committee on Human Research (#11–05539 and #11–05995) and the Uganda National Council for Science and Technology (#HS-978 and #HS-1019).

Samples

Sera were originally collected as part of a comprehensive longitudinal surveillance study conducted in three sub-counties in Uganda (Walukuba, Jinja District; Kihihi, Kanungu District, and Nagongera, Tororo). The study design and methods have been previously reported and are described in detail elsewhere [44]. A sub-selection of samples (n = 899) was made from individuals across a breadth of recorded clinical episodes of malaria to ensure a range of sero-reactivity.

Protein targets

Purified protein expression

Recombinant proteins were generated and expressed in Escherichia coli as glutathione S-transferase (GST)-tagged fusion proteins using previously described methods: PfMSP1-19 [45]; MSP1 block 2 [46]; ACS5, ETRAMP4 & HSP40 [19]; ETRAMP5 [19, 47]; EBA181 [48]; MSP4 [49]; MSP5 [50]; MSP7 [51]; and GAMA [52]. The exception to this was PfAMA1, which was expressed as a histidine tagged protein in Pichia pastoris [53]. Purification of the expressed proteins was performed using affinity chromatography (Glutathione Sepharose 4B (GE Healthcare Life Sciences) or HisPur Ni-NTA (Invitrogen) resins for GST and His tagged proteins, respectively). Protein concentration was assessed using the Bradford protein assay, with quality, and purity assessed by resolution on a 4–20% gradient SDS-PAGE.

IVTT protein expression

An IVTT system was used to express proteins of interest as previously described [15]. Briefly, Plasmodium falciparum DNA (3D7 isolate) coding sequences were PCR-amplified and cloned into T7 expression vectors via homologous recombination. Target sequences were expressed at 21°C for 16h in E. coli-based, cell-free transcription/translation reactions, and products were printed onto arrays as un-purified, whole reaction mixtures.

Overview of compared IVTT and purified protein antigens

We assessed antibody responses to protein targets mapping to eleven antigens (i.e. distinct gene products), each represented on the array by at least one IVTT and one purified protein target. Full details are in and . The number of purified protein targets varied according to availability, while the number of IVTT targets was dependent on the exon composition of each the gene sequence; multiple exon sequences were expressed as multiple protein targets based on exon delineation. Similarly, single exon gene sequences were generally expressed as a single protein. As a result, of the 11 antigens investigated, 8 were represented by >1 IVTT or purified protein target; 5 had >1 IVTT protein target (EBA181, HSP40, MSP1, MSP4 and MSP5) and 5 had >1 purified protein target (ACS5, ETRAMP4, ETRAMP5, HSP40 and MSP1). Near identical IVTT proteins (1 terminal amino acid difference in length) were produced independently and printed in parallel for two antigens: MSP4 and MSP5 as expression controls. Sequence information used in the design and expression of the purified E.coli proteins were generally smaller than the equivalent proteins expressed in the IVTT cell-free systems. This was done to limit the sequence length to below 1kb as expression of proteins larger that 1kb in E.coli can contribute to poor or failed expression yields [42, 43]. Truncation of target sequences was based on in silico mapping of each protein sequence to focus on regions of predicted immunogenicity based on the in silico analysis. Empty GST vectors were expressed and the purified GST used in background correction for proteins with this tag. His-tag vector was not expressed as it has proven impossible to express and purify the 6xhistidine tag in isolation.

Protein microarray

Prior to printing, Tween 20 was added to purified proteins to yield a final concentration of 0.001% Tween 20. Arrays were printed onto nitrocellulose-coated slides (AVID, Grace Bio-Labs, Inc., Bend, OR, USA) using an Omni Grid Accent microarray printer (Digilabs, Inc., Marlborough, MA, USA). Alongside proteins of interest, buffer (PBS) and no-DNA (empty T7 vector reactions) were included as controls to allow for background normalisation of purified and IVTT proteins respectively.

Sample probing

For analysis of antibody reactivity on the protein microarray, serum samples were diluted 1:200 in a 3 mg mL-1 E. coli lysate solution in protein arraying buffer (Maine Manufacturing, Sanford, ME, USA) and incubated at room temperature for 30 min. Arrays were rehydrated in blocking buffer for 30 min. Blocking buffer was removed, and arrays were probed with pre-incubated serum samples using sealed, fitted slide chambers to ensure no cross-contamination of sample between pads. Chips were incubated overnight at 4°C with agitation. Arrays were washed five times with TBS-0.05% Tween 20, followed by incubation with biotin-conjugated goat anti-human IgG (Jackson ImmunoResearch, West Grove, PA, USA) diluted 1:200 in blocking buffer at room temperature. Arrays were washed three times with TBS-0.05% Tween 20, followed by incubation with streptavidin-conjugated SureLight P-3 (Columbia Biosciences, Frederick, MD, USA) at room temperature protected from light. Arrays were washed three times with TBS-0.05% Tween 20, three times with TBS, and once with water. Arrays were air dried by centrifugation at 500 x g for 5 min and scanned on a GenePix 4300A High-Resolution Microarray Scanner (Molecular Devices, Sunnyvale, CA, USA). Target and background intensities were measured using an annotated grid file (.GAL).

Data normalisation

Microarray spot foreground and local background fluorescence data were imported into R (Foundation for Statistical Computing, Vienna, Austria) for correction, normalisation and analysis. Local background intensities were subtracted from foreground using the backgroundCorrect function of the limma package [54]. The backgroundCorrect function was then further applied to GST-tagged purified proteins, whereby background-corrected GST fluorescence was subtracted from background-corrected target fluorescence to account for any GST-specific reactivity in samples. All data were then Log2 transformed and the mean signal intensity of buffer and no-DNA control spots were subtracted from purified and IVTT proteins respectively to give a relative measure of reactivity to targets over background () [20].

Results

summarises the purified and IVTT protein targets for each antigen, with further detail in . In brief, we assessed IgG antibody responses to 35 antigenic targets, derived from 11 well-characterised P. falciparum protein antigens (distinct gene products). Each antigen was represented by at least one IVTT and one purified protein target.

Magnitude of responses between expression systems

The magnitude of response to all protein targets was compared by antigen to evaluate differences in seroreactivity between IVTT derived and purified protein targets. As expected, responses varied significantly between antigens and between the protein targets mapping to each antigen.

Mean responses to all targets were compared by expression system () revealing a greater range of response to purified proteins (IQR Log2MFI = 3.88–6.40) than IVTT proteins (IQR Log2MFI 0.46–1.68), and a greater magnitude of response to purified than IVTT targets (p = <0.001). Similarly, the range and median intensity of individual antibody responses was found to be greater for purified proteins than their IVTT counterparts (e.g. AMA1—IVTT_1, median [IQR] Log2MFI = 1.66 [0.80–2.53]; Pure_1, median [IQR] Log2MFI = 7.92 [6.16–8.52]) for all targets (p = <0.001) except MSP1 Pure_2, which more closely reflected the level of reactivity to the two MSP1 IVTT targets ().

Mean magnitude of antibody responses to targets.

The mean magnitude of response of each protein target stratified by expression system, presented with median and interquartile range of all mean responses.

Magnitude and range of response to IVTT and purified proteins.

All sample responses (n = 899) to all protein targets grouped by antigen, presented with median and interquartile range.

Correlation of responses between antigen matched targets

Considering all at least partially sequence matched IVTT and purified protein targets (i.e. excluding pairwise comparisons where purified protein sequence were completely non-overlapping with IVTT sequence for the same antigen) there was no evidence for a general correlation in mean response between expression platforms (Spearman’s rho (rs) = 0.279, p = 0.23). Antibody responses to all protein targets for each antigen were therefore compared individually (representative example in and all antigens in ). This allowed for comparison between sequence matching IVTT and purified protein targets (e.g. HSP40 IVTT 2 vs. HSP40 Pure 2), non-matching IVTT and purified protein targets (e.g. HSP40 IVTT 2 vs. HSP40 Pure 1), and matching or non-matching targets produced in the same system (e.g. HSP40 IVTT 1 vs. HSP40 IVTT 2). Correlations were highly variable (rs = 1.00 to -0.045) though all but one (GAMA; rs = -0.045, p = 0.17) demonstrated a degree of positive, if not always statistically significant, association.

Correlation of antibody responses and sequence mapping.

A representative example correlogram of multiple antigen-matched targets (left). Spearman’s rank correlation reported (rs) and increasing blue colour scale indicates relative strength of correlation based on calculated correlations for all proteins included in this analysis. Protein schematic (right) represents amino-acid aligned representation of IVTT (green) and purified (orange) proteins to the full-length native protein (grey). Proteins in the correlogram and schematic are correspondingly aligned. Corresponding axes are adjacently below or to the left of each protein.

Multiple IVTT targets were produced for EBA181, HSP40, MSP1, MSP4 and MSP5. For all other than MSP5, non-sequence matching IVTTs were produced; correlation co-efficient for these targets were between 0.37 and 0.73 (). For EBA181 and MSP1, IVTT targets overlap by 17 amino acids–equivalent to a small peptide in terminal regions unlikely to cover immunogenic epitopes. As such, these targets were considered non-overlapping. For MSP4 and MSP5, duplicate IVTT protein products were generated for each gene, with each duplicate protein identical to the other except for the omission of one [N- or C-] terminal amino acid. These respective targets resulted in near perfect correlation of antibody responses (MSP4 rs = 1.00, p = <0.001; MSP5 rs = 0.94, p = <0.001). Multiple purified protein targets were produced for ACS5, ETRAMP4, ETRAMP5, HSP40 and MSP1—none of which overlap. Correlation between these purified protein targets in each antigen varied between 0.31 and 0.59 ().

For the 8 antigens with >1 IVTT or purified protein target, the greatest level of correlation was found between an IVTT and purified target in 4/8 instances; between two IVTT targets (IVTT-IVTT) in 3/8 instances; and between two purified targets (purified-purified) in 1/8 instances (). Comparing correlations between antigen-matched IVTT and purified proteins only, overlapping targets correlate more highly than non-overlapping targets. Sample sizes were too low to test the significance of this trend within antigens ().

Spearman’s rho correlation coefficients between antigen-matched IVTT and purified proteins.

Targets with overlapping amino acid sequences are indicated by closed circles, compared to non-overlapping sequences indicated by open circles.

Discussion

Protein microarrays are a practical approach to the serological screening of large numbers of putative malaria antigen biomarkers. The throughput and flexibility of the microarray platform presents an opportunity to interrogate malarial antibody responses at a scale far exceeding traditional mono- or multiplex approaches, agnostic of predicted immunological targets. Here we have evaluated matched antigenic targets produced using two E. coli-based expression techniques–in vitro transcription/translation (IVTT), and purified, whole-cell recombinants–in the context of a protein microarray. We found that the magnitude of antibody responses to purified protein targets was generally higher than for their IVTT counterparts, and that correlation between protein target pairs at the individual serum sample level was variable and related to degree of sequence homogeneity between targets. Our findings warn against direct comparisons of microarray data from proteins produced in different expression platforms without careful cross-validation of sequences and allelic types. However, our data do provide support for the use of both IVTT and purified protein microarray platforms in the context of early-stage antigen biomarker identification to feed into experimental pipelines where candidate proteins may be interrogated by methods providing higher resolution analysis.

In building this study, we predicted that the magnitude of responses to IVTT products–which tended to be longer, often representing single exon sequences and therefore potentially containing more epitopes–would be greater than purified targets truncated based on species-specificity or domain boundaries which potentially represented fewer epitopes. Contrary to this prediction, we found that purified proteins captured a greater range and magnitude of responses (Purified, IQR Log2MFI = 3.88–6.40; IVTT, IQR Log2MFI 0.46–1.68; p = <0.001). The greater level of reactivity to purified targets may relate to differences in the amount of protein deposited on the array, where consistent and defined amounts of purified protein are spotted in contrast to the unquantified, and likely variable IVTT products. These findings recommend a degree of caution in interpretation of array data from two different platforms, for example: MSP5 showed the second highest mean MFI for any purified protein, but showed among the lowest mean MFI of any IVTT protein.

In addition to differences in the magnitude of mean responses to targets stratified by expression system, we observed a greater range of individual sample responses, stratified by antigen, to purified proteins than in sequence matched IVTT-expressed targets (e.g. AMA1—IVTT_1, median [IQR] Log2MFI = 1.66 [0.80–2.53]; Pure_1 (Pichia pastoris produced), median [IQR] Log2MFI = 7.92 [6.16–8.52]; p = <0.001). The P. pastoris AMA1 was included as a control for the evaluation of the production of a conformational protein. AMA1 is a complex structure comprised of three domains defined by three disulphide bonds. Production of AMA1 in P. pastoris has been fully characterised in terms of correct folding of the purified protein [53, 55] and this observation is likely a reflection of antibody reactivity to correctly folded (P. pastoris) and incorrectly folded AMA1 (IVTT). We acknowledge that a lack of correct folding in other purified and IVTT products may impact on epitope recognition by antibodies raised to native protein during infection. However, human antibody responses are composed of a polyclonal response to each antigen, which will include both confirmation and linear epitopes. Whilst questions remain about the appropriateness of using unfolded protein fragments in serological screens, such reagents remain the most widely utilised and efficient approach in this context at present.

Considering all antigenic targets together, we found no evidence of correlation in mean reactivity to sequence matched targets between expression systems (rs = 0.28, p = 0.23). In the context of this study, this was not unexpected taking into account the differences observed in magnitude of response between IVTT and purified proteins, and that the length of native protein sequence coverage between IVTT and purified targets was highly variable. More broadly, it is perhaps less reassuring that matched targets derived from different expression systems lack more obvious relationships in antibody response than have been demonstrated in other studies [17, 56], though Kobayashi et al. report relatively similar results for a smaller number of targets expressed in E. coli (purified proteins) and IVTT systems specifically [30]. It is likely that protein concentration disparities between the two approaches are one of the drivers of this heterogeneity. However, without attempting to quantify the exact amount of protein generated in the small volume of IVTT reactions we are unable to address this here. Although in this current study targets grouped by antigen displayed highly variable correlations of response, it is encouraging that sequence matched proteins did generally display stronger correlations of response than non-sequence matched targets. Further, this may indicate the importance of capturing specific epitopes within expression sequences when producing antigens by either expression method.

Despite the lack of a clearly defined relationship between antigen-matched targets from the evaluated expression systems, we remain confident that microarrays utilising IVTT or purified recombinant proteins are able to produce compelling and biologically relevant data. Indeed, our data show age-dependent trends in antibody responses (typical of highly endemic populations) [1-3] irrespective of expression system (), lending weight to the applicability of either methodology in serological assays [57-70].

The IVTT system lends itself to microarray applications, as vast numbers of proteins, or even entire proteomes, may be produced at scale relatively quickly. However, for application to serology there is concern that expressed proteins are not quantified before printing, and that expression levels of product may vary considerably; product yield in bacterial-based IVTT systems is generally considered to be lower (typically ~1 mg mL-1 or less) though higher protein yields have been reported [71, 72]. This has been shown to be due to an inherent heterogeneity with IVTT components, although this weakness is an area of active research [26, 73]. Similarly, it is important to acknowledge that the un-purified nature of printed reaction mixtures may mask, or otherwise adversely affect, the detection of antibody reactivity in a sample; Davies et al. report IVTT reaction compositions of 99% E. coli lysate to 1% target protein [15], though this will vary considerably, at scale, in practice.

In contrast to IVTT-based microarrays, printing purified protein allows a highly quantifiable approach to be taken. Affinity purification and dialysis of expression products substantially reduces the risk of background reactivity to bacterial components, and the simple determination of target protein concentrations allows defined quantities of product to be spotted, providing much greater confidence when comparing reactivity between targets. However, these advantages come at a substantial cost; the need for in silico analysis to design vectors, transfection procedures, expression and purification drastically slows the rate at which putative targets can be produced and screened. Shorter, epitope specific sequences may in theory be transposed from IVTT systems with a view to generating more granular serological screens, though we accept that truncated protein targets will in some cases favour linear B cell epitopes, while missing conformational epitopes. However, for measuring exposure to infection there is less importance on the targeting of confirmation epitopes than would be required for protective epitopes [57].

The primary benefit of the microarray platform is the ability to screen orders of magnitude more targets simultaneously than more standard serological assays. Our analysis shows that both IVTT and purified proteins can be successfully used to capture malarial protein-antigen specific antibody responses on a protein microarray. Although correlations of response between expression systems are not as strong as may have been expected, a number of acknowledged technical differences in the methods of protein production may account for this finding. In addition to the E. coli in vivo and IVTT systems utilised here, high-throughput wheat germ cell free systems have been successfully used to conduct large scale serological screens of putative antigen biomarkers [74, 75], alongside chemically synthesised peptide arrays [57, 62]. High-throughput mammalian and baculovirus expression systems have also been pioneered for the production of recombinant proteins [36, 76]. Differences in expression efficiency and the homology to native epitopes achieved by the assortment of available approaches likely have considerable impact on the capture of antibody from sample. This variability should be accounted for both in terms of choosing an experimental approach and comparative analysis between different methods. We suggest that further investigation of differences in seroreactivity to sequence-matched proteins derived from contrasting expression systems is needed to shed light on the parity between such data that is already widely published. It should also be noted that it is unlikely that any single expression platform will satisfy the demands of all recombinant expression projects due to varying importance such as protein folding, proteins activity (e.g. enzymes) and glycosylation. In addition, E. coli expression has the advantage of low cost, flexibility and easy scale-up.

Considering the data presented here more broadly, observed trends lend support to the utilisation of both IVTT and purified arrays depending on the objectives and context of hypotheses to be investigated. The strengths and weaknesses of each expression system should dictate the chosen approach on a case-by-case basis. For example, very high-density proteome level screening to identify ‘shortlists’ of candidate markers based on binary categorisation of seropositivity may be best achieved using IVTT systems. In contrast, smaller numbers of ‘shortlisted’ targets expressed as purified proteins may allow for more nuanced characterisation of antibody responses on a more continuous scale. As already described, the key limitation in the production of purified recombinants in our current expression pipeline is throughput. The adaption of our methods to increase the capacity of protein production would improve our ability to more widely mine the biomarker information derived from the IVTT platform. As such, we are currently exploring a number of existing approaches to address this methodological bottleneck [38, 77].

In summary, the IVTT protein microarray approach has proven to be a powerful, high-throughput, biomarker discovery platform with applicability across a range of infectious diseases. When combined with a cheap, scalable and flexible protein expression platform such as the E. coli in vivo expression platform we have the ability to mine potential diagnostic and vaccine related targets.

Data normalisation processes for IVTT and purified protein spots.

After local background correction using the backgroundCorrect function from the limma package, purified protein spots were additionally corrected for possible GST reactivity by subtracting GST reactivity using the same function. After Log2 transformation, IVTT and purified proteins were normalised to background control spots of empty T7 vector and PBS buffer control spots respectively.

(PDF)

Click here for additional data file.

Correlogram of multiple antigen-matched targets (left). Spearman’s rank correlation reported (rs) and increasing blue colour scale indicates relative strength of correlation based on calculated correlations for all proteins included in this analysis. Protein schematic (right) represents amino-acid aligned representation of IVTT (green) and purified (orange) proteins to the full-length native protein (grey). Proteins in the correlogram and schematic are correspondingly aligned.

(PDF)

Click here for additional data file.

Magnitude and range of response to IVTT and purified proteins, stratified by age.

All sample responses (n = 899) to all protein targets grouped by antigen, presented with median and interquartile range.

(PDF)

Click here for additional data file.

Detail of expressed protein targets.

A key to the simplified nomenclature used for specific proteins in text is provided.

(XLSX)

Click here for additional data file.

Correlation coefficient results for all protein pairs.

Protein targets are grouped by antigen, and all possible combinations within each antigen group are shown.

(XLSX)

Click here for additional data file.

22 Dec 2021

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Takafumi Tsuboi

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and

https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. Thank you for stating the following in the Acknowledgments Section of your manuscript:

"We thank all the study participants who participated in the original PRISM study (Program for Resistance, Immunology, Surveillance, and Modelling of Malaria in Uganda; East Africa ICEMR) from which the test samples were taken. The authors give thanks to James Beeson for the provision of purified recombinant protein EBA140 RIII-V, Ross Coppel for the provision of MSP4 and MSP5, and Tony Holder for the provision of GAMA and MSP7. Kevin K.A.Tetteh was supported by a Bloomsbury SET Award (Innovation Fellowship to KKAT; BSA14) under the UKRI Connecting Capabilities Fund (CCF)."

We note that you have provided funding information that is not currently declared in your Funding Statement. However, funding information should not appear in the Acknowledgments section or other areas of your manuscript. We will only publish funding information present in the Funding Statement section of the online submission form.

Please remove any funding-related text from the manuscript and let us know how you would like to update your Funding Statement. Currently, your Funding Statement reads as follows:

"This work was supported by funding from the Global Good Fund I, LLC. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript."

Please include your amended statements within your cover letter; we will change the online submission form on your behalf.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: No

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: In this study, Oulton et al compared the reactivity of recombinant proteins synthesized using different platforms. Specifically, they compared human sera reactivity to purified or unpurified recombinant proteins synthesized in E . coli /Pichia pastoris or an in vitro transcription/ translation (IVTT) platform, respectively.

Although the study is important since it attempts to address an existing gap, it has several weaknesses. First, it fails to acknowledge the great progress in antigen discovery that have been made in the last decade rather basing the current work on studies conducted more than a decade. The authors only briefly mention the high-throughput studies in the discussion. The introduction section should updated to clearly capture the current status as reported in more recent references. Second, key data is missing or the design didn’t not fully answer the researchers questions. The purity of the proteins is unknown since no data was presented. Moreover, it’s impossible to tell whether the low signals observed with the IVTT expressed proteins was due to failure of expression or was due to poor quality of the proteins. Since the IVTT expression system is scalable, at least several purified IVTT proteins should have been included as controls.

Specific comments

Line 46-47 is not clear what “the rate of IVTT” means. This line needs editing.

For clarity, the authors should – at first appearance- clearly state that they compared the reactivity of recombinant proteins expressed in E. coli or E. coli based IVTT proteins expression system. As it is now, it not clear how the “purified proteins” were synthesized.

Since this study is about assessing the effect of purification on proteins, the SDS-PAGE images, of the study proteins should be made available to the readers.

Line 110-112: “Purified protein targets were smaller than their IVTT counterparts; purified protein targets are produced as fragments of the full-length protein, with the aim of capturing antibodies specific to predicted epitopes based on in silico analysis”. Does it mean that the IVTT proteins were not detecting antibodies specific to predicted epitopes? Please correct.

Did the empty vectors express GST or the tags was expressed separately for background GST fluorescence correction? How were the His-tagged proteins normalized?

Line 188-190: If the amino acid labeling is correct, the difference was only at the the c-terminal amino acid but not N-terminal.

I note that this study was part of a very comprehensive longitudinal study in Uganda. Can the authors confirm that only one Ugandan researcher should be included as a co-authors?

Reviewer #2: The study by Oulton, et al, describes the ability of IgG antibodies to bind to Pf antigens whether produced by an IVTT system or through recombinant whole-cell system (and purified). This is an important question for the malaria serological community, as high-throughput fishing expeditions can lead to downstream decision making, but results are rarely compared to the selected produced product. The data shows, not surprisingly, that the recombinant antigens have a better capacity for IgG capture compared to comparable targets produced by IVTT.

Major comments

A major criticism of this work is how the authors compare the ability of IgG capture only through the microarray platform on nitrocellulose slides. The reviewer would be certain that these samples (or a subset) have also been assayed by a separate immunoassay platform such as ELISA or multiplex bead assay. Comparison of signal intensity for single persons’ samples between the microarray and one of these standard IgG detection assays would be an important bridge for the laboratorian to translate these findings.

The concept of background signal due to a crude IVTT system with a lot of E. coli protein in it is brought up a few times, though no data is shown for what this background was in the authors’ hands. Was the background MFI of the IVTT on the microarray considerably higher than for the purified antigens? Did this lead to an overall reduced signal IVTT targets after normalization?

Minor comments

Line 25-26: “lack of a clearly-defined relationship between…” confusing text here, should consider re-wording

Line 46-47: “rates of IVTT..” rates meaning hours expended, time for each target, other? Please provide a little more detail here regarding ‘rates’

Line 49: write out E. coli first time used

Line 53: Sentence starting with ‘However’ can be removed

Line 63: say why the production of complex conformational proteins “can sometimes be a challenge”

Table 1: “Description of P. falciparum antigens and…”

Figure 1: either in this panel (or a new one) it would actually be more informative for the reader to have a line connecting the IVTT MFI value to the ‘purified’ value for each individual target. In that way for a single target, the relative log increase in signal is clearly shown.

Line 214-217: the authors have argued previously that IVTT is more appropriate for the broad screening of biomarker identification. In fact, a recombinant and purified antigen is no longer a ‘candidate’, but there would have been a reason for going to all the trouble to create it in a cell system.

Throughout: the authors should note that the lack of correlation for many of their targets is due to the fact that the purified antigen is maxing out the MFI signal whereas is seems IVTT targets never do this.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

19 Jul 2022

Reviewer #1: In this study, Oulton et al compared the reactivity of recombinant proteins synthesized using different platforms. Specifically, they compared human sera reactivity to purified or unpurified recombinant proteins synthesized in E . coli /Pichia pastoris or an in vitro transcription/ translation (IVTT) platform, respectively.

Although the study is important since it attempts to address an existing gap, it has several weaknesses. First, it fails to acknowledge the great progress in antigen discovery that have been made in the last decade rather basing the current work on studies conducted more than a decade. The authors only briefly mention the high-throughput studies in the discussion. The introduction section should updated to clearly capture the current status as reported in more recent references.

Author response:

We thank the reviewer for this comment. This omission was not intentional. Additional text and supporting references have been included into the introduction to address this gap as follows:

Line 46-57

“Cell-free synthesis (CFS) is a technique first established over 50 years ago as a means to dissect the molecular mechanisms around protein expression. More recently, the technique has been used as a high throughput expression platform to explore a number of diverse biological processes (Dopp et al Synthesis and systems biology 4, PMID: 31750411; Harini et al. Current Opinions in Biotechnology 2022, PMID: 35247659). At its simplest, the approach utilises the crude extract containing the transcription and translation machinery from the cell, performing the process of protein expression without the constraints of the cell. This allows a wide variety of proteins to be expressed including those that would be deemed toxic if expression was attempted within the confines of the cell membrane (Silverman et al Nat Rev Gen 2019, PMID: 31780816). CFS systems based on Escherichia coli (E.coli) are among the most widely used of the IVTT systems (Harini et al Current Opinions in Biotechnology 2022, PMID: 35247659) and have helped to transform the narrative around a number of areas including biomarker discovery for infectious diseases (Davies DH et al. Proc Natl Acad Sci U S A. 2005 Jan 18;102(3):547-52. PMID: 15647345; Venkatesh A et al. Methods Mol Biol. 2021;2344:139-150. PMID: 34115357; Kobayashi T et al. mSphere. 2019 Mar 27;4(2):e00061-19. PMID: 30918058). Despite the widespread uptake of the approach there remain some issues with the technique. This includes significant heterogeneity of expression, leading some research groups to describe the mechanisms of the process as a “black box”.”

Second, key data is missing or the design didn’t not fully answer the researchers questions. The purity of the proteins is unknown since no data was presented. Moreover, it’s impossible to tell whether the low signals observed with the IVTT expressed proteins was due to failure of expression or was due to poor quality of the proteins. Since the IVTT expression system is scalable, at least several purified IVTT proteins should have been included as controls.

Author response:

We thank the reviewer for their comments. The data is not missing but is in fact a recognised fundamental weakness in the IVTT platform. The low signals described in this study and covered extensively in the literature is a known limitation of the system. The high throughput nature of the system is somewhat tempered by the high level of heterogeneity in protein production that is inherent with the system. The IVTT microarray approach is more of an ‘opt in’ where positives can be further validated using other approaches. This circumvents the heterogeneity which can result in detectable signal in one expression run vs no signal in another. We have included additional text highlighting the state-of-the-art with regards to the IVTT platform including the following references in the introduction (Dopp et al Synthesis and systems biology 4, PMID: 31750411; Harini et al. Current Opinions in Biotechnology 2022, PMID: 35247659).

Although the system is scalable, in terms of volume of reaction mix/culture, this does not circumvent the known heterogeneity that is inherent within the system. This is an issue that groups are trying to address as an end goal. Large-scale systems do exist within the biotechnology industry, but these are not generally accessible to researchers due to space considerations and cost to run (Silverman et al Nat Rev Gen 2019, PMID: 31780816)

Relevant controls are built into the array and are described under heading “Protein microarray”.

Refer to inserted text as above (Line 46-57)

Specific comments

Line 46-47 is not clear what “the rate of IVTT” means. This line needs editing.

For clarity, the authors should – at first appearance- clearly state that they compared the reactivity of recombinant proteins expressed in E. coli or E. coli based IVTT proteins expression system. As it is now, it not clear how the “purified proteins” were synthesized.

Since this study is about assessing the effect of purification on proteins, the SDS-PAGE images, of the study proteins should be made available to the readers.

Author response:

The line previously at 46-47 has been deleted and replaced with more detailed text as highlighted above for Line 46-57

We agree with the reviewer regarding the difference in expression systems and have reworked the text for clarity as follows:

For clarity proteins produced using the IVTT system will simply be referred to as IVTT proteins, and those produced by conventional E.coli expression will be referred to as purified proteins. The text has been edited as follows:

Line 77-79

“For clarity proteins produced using the IVTT system will simply be referred to as IVTT proteins, and those produced by conventional E.coli expression will be referred to as purified proteins.”

The purified proteins described here have been fully described elsewhere and so were not duplicated here.

Line 110-112: “Purified protein targets were smaller than their IVTT counterparts; purified protein targets are produced as fragments of the full-length protein, with the aim of capturing antibodies specific to predicted epitopes based on in silico analysis”. Does it mean that the IVTT proteins were not detecting antibodies specific to predicted epitopes? Please correct.

Did the empty vectors express GST or the tags was expressed separately for background GST fluorescence correction? How were the His-tagged proteins normalized?

Author response:

This section has now been clarified as follows:

Line 125-1323

“Sequence information used in the design and expression of the purified E.coli proteins were generally smaller than the equivalent proteins expressed in the IVTT cell-free systems. This was done to limit the sequence length to below 1kb as expression of proteins larger that 1kb in E.coli can contribute to poor or failed expression yields (Vedadi M et al. Mol Biochem Parasitol. 2007 Jan;151(1):100-10. PMID: 17125854; Mehlin C et al. Mol Biochem Parasitol. 2006 Aug;148(2):144-60. PMID: 16644028). Truncation of target sequences was based on in silico mapping of each protein sequence to focus on regions of predicted immunogenicity based on the in silico analysis. Empty GST vectors were expressed and the purified GST used in background correction for proteins with this tag. His-tag vector was not expressed as it has proven impossible to express and purify the 6xhistidine tag in isolation.”

Line 188-190: If the amino acid labeling is correct, the difference was only at the the c-terminal amino acid but not N-terminal.

Author response:

As part of the design of the protein targets for expression, the signal peptide and any transmembrane proteins are removed from the sequences, which is standard practice with expression of target proteins in E.coli. Meaning that some edits can be based around the signal peptide only, resulting in minor truncations of the target sequences. Inclusion of the signal peptide and transmembrane domains are key factors in the poor or failed expression of recombinant proteins.

Supplementary Table S1 provides detailed information on each target to help illustrate this.

I note that this study was part of a very comprehensive longitudinal study in Uganda. Can the authors confirm that only one Ugandan researcher should be included as a co-authors?

Author response:

We thank the reviewer for this comment. As the lead of our Equity and Diversity committee and one of the key drivers in pushing for better equity in research I am encouraged by this question and thank the reviewer for raising this. The inclusion of authors into this manuscript was conducted in an open and fair manner, with final decision for author inclusion from Uganda resting with Dr Ssewanyana (Director of Laboratory Services Uganda National Health Laboratory Services). Ultimately, the inclusion of Ugandan staff relevant to the project rested with him, and this is a decision all authors supported. In addition, there have been over 246 publications from PRISM 1 and 2 for which there has been equitable contribution in terms of authorship across the collaborative partners (https://www.niaid.nih.gov/research/east-africa-international-center-excellence-malaria-research).

Reviewer #2: The study by Oulton, et al, describes the ability of IgG antibodies to bind to Pf antigens whether produced by an IVTT system or through recombinant whole-cell system (and purified). This is an important question for the malaria serological community, as high-throughput fishing expeditions can lead to downstream decision making, but results are rarely compared to the selected produced product. The data shows, not surprisingly, that the recombinant antigens have a better capacity for IgG capture compared to comparable targets produced by IVTT.

Major comments

A major criticism of this work is how the authors compare the ability of IgG capture only through the microarray platform on nitrocellulose slides. The reviewer would be certain that these samples (or a subset) have also been assayed by a separate immunoassay platform such as ELISA or multiplex bead assay. Comparison of signal intensity for single persons’ samples between the microarray and one of these standard IgG detection assays would be an important bridge for the laboratorian to translate these findings.

The concept of background signal due to a crude IVTT system with a lot of E. coli protein in it is brought up a few times, though no data is shown for what this background was in the authors’ hands. Was the background MFI of the IVTT on the microarray considerably higher than for the purified antigens? Did this lead to an overall reduced signal IVTT targets after normalization?

Author response:

The comparison of IgG in this study using the protein microarray multiplex assay was performed based on prior research from a number of groups including ours using the monoplex ELISA to assess antibody responses to infection based primarily on IgG but also targeting subclass and isotype responses (van den Hoogen LL et al. Front Microbiol. 2019 Jan 16;9:3300; Kamuyu G et al. Front Immunol. 2018 Dec 11;9:2866. PMID: 30619257; King CL et al. Am J Trop Med Hyg. 2015 Sep;93(3 Suppl):16-27. PMID: 26259938; Helb DA et al. Proc Natl Acad Sci U S A. 2015 Aug 11;112(32):E4438-47. PMID: 26216993). Many of the key markers of seroincidence have also been extensively assessed using the Luminex multiplex bead-based array (Wu L et al. EClinicalMedicine. 2022 Feb 14;44:101272. PMID: 35198913; Achan J et al. Lancet Microbe. 2022 Jan;3(1):e62-e71. PMID: 34723228). Comparisons between the microarray and ELISA have already been (Ondigo BN et al. Malar J. 2012 Dec 21;11:427. PMID: 23259607) and so was not repeated here.

Primarily, the IVTT is a uniquely high-throughput platform capable of generating 100’s to 1000’s of targets in parallel. However, levels of expression can be highly variable. Our study was intended to compare the E.coli IVTT vs the E.coli purified protein responses, specifically within the context of microarray analysis. This study was intended to support the assertion that biomarker discovery can feed into experimental pipelines that can be interrogated using high resolution methods.

Data processing for the IVTT microarray has been presented extensively elsewhere. No. Background MFI for the IVTT did not show higher background responses that their purified antigen counterparts. As such there was not bias in terms of an overall reduced signal for the IVTT proteins. Differences in signal intensity were due to the inherent heterogeneity of expression with the IVTT platform as described earlier.

Minor comments

Line 25-26: “lack of a clearly-defined relationship between…” confusing text here, should consider re-wording.

Author response:

We agree with the reviewer and have edited the sentence for clarity from:

“Despite the lack of a clearly defined relationship between antigen-matched targets produced in each expression system, our data indicate that protein microarrays produced using either method can be used confidently, in a context dependent manner, though care should be taken when comparing data derived from contrasting approaches.”

To,

Line 25-29

“Despite the lack of a clear correlation between antigen-matched recombinant proteins from each expression system, our data indicates that protein microarrays produced using either method can be used confidently, in a context dependent manner, though care should be taken when comparing data derived from contrasting approaches.”

Line 46-47: “rates of IVTT..” rates meaning hours expended, time for each target, other? Please provide a little more detail here regarding ‘rates’

Author response:

This section has been reworked for improved clarity as follows:

The line at 46-47 has been deleted and replaced with more detailed text as follows:

Line 46-57

“Cell-free synthesis (CFS) is a technique first established over 50 years ago as a means to dissect the molecular mechanisms around protein expression. More recently, the technique has been used as a high throughput expression platform to explore a number of diverse biological processes (Dopp et al Synthesis and systems biology 4, PMID: 31750411; Harini et al. Current Opinions in Biotechnology 2022, PMID: 35247659). At its simplest, the approach utilises the crude extract containing the transcription and translation machinery from the cell, performing the process of protein expression without the constraints of the cell. This allows a wide variety of proteins to be expressed including those that would be deemed toxic if expression was attempted within the confines of the cell membrane (Silverman et al Nat Rev Gen 2019, PMID: 31780816). CFS systems based on Escherichia coli (E.coli) are among the most widely used of the IVTT systems (Harini et al Current Opinions in Biotechnology 2022, PMID: 35247659) and have helped to transform the narrative around a number of areas including biomarker discovery for infectious diseases (Davies DH et al. Proc Natl Acad Sci U S A. 2005 Jan 18;102(3):547-52. PMID: 15647345; Venkatesh A et al. Methods Mol Biol. 2021;2344:139-150. PMID: 34115357; Kobayashi T et al. mSphere. 2019 Mar 27;4(2):e00061-19. PMID: 30918058). Despite the widespread uptake of the approach there remain some issues around the technique. This includes significant heterogeneity of expression, leading some research groups to describe the mechanisms of the process as a “black box”. “

Line 49: write out E. coli first time used

Author response:

Line 53

We have made this change.

Line 53: Sentence starting with ‘However’ can be removed

Author response:

Line 63

We have removed, ”However”.

Line 63: say why the production of complex conformational proteins “can sometimes be a challenge”

Author response:

Expression of recombinant proteins is challenging. This is in part due to the expression of proteins foreign to the bacteria, the speed at which bacteria express proteins, only partially mitigated with a reduction in expression temperature; and the lack of essential molecular chaperones to aid correct folding/refolding of proteins (Francis and Page Curr Protoc Protein Sci. 2010 Aug; 61(1): 5241–52429.)

The following sentence has been included for clarity:

Line 73-76

“These challenges are in part due to the expression of proteins foreign to the bacteria, the speed at which bacteria express proteins, only partially mitigated with a reduction in expression temperature; and the lack of essential molecular chaperones to aid correct folding/refolding of proteins (Francis and Page Curr Protoc Protein Sci. 2010 Aug; 61(1): 5241–52429; Vedadi M et al. Mol Biochem Parasitol. 2007 Jan;151(1):100-10. PMID: 17125854; Mehlin C et al. Mol Biochem Parasitol. 2006 Aug;148(2):144-60. PMID: 16644028).

Table 1: “Description of P. falciparum antigens and…”

Author response:

The recommended edit has been made

Figure 1: either in this panel (or a new one) it would actually be more informative for the reader to have a line connecting the IVTT MFI value to the ‘purified’ value for each individual target. In that way for a single target, the relative log increase in signal is clearly shown.

Author response:

We appreciate the reviewer’s comments. However, the recommended edits would lead to a less transparent figure. In figure 1, we aim to demonstrate differences in reactivity between proteins produced by the two expression systems overall, rather than between specific antigens. Antigens were colour coded to allow for a basic comparison of reactivity between antigen matched targets, though this is not the main purpose of this figure. We would instead like to draw the reviewer’s attention to Figure 2, which highlights both the magnitude and range of responses of the IVTT proteins compared to the purified proteins.

Similarly, to Figure 3 and supplementary Figure 2 which show the correlation of responses between the IVTT and the purified proteins.

We believe that these figures capture what the reviewer was trying to highlight.

Line 214-217: the authors have argued previously that IVTT is more appropriate for the broad screening of biomarker identification. In fact, a recombinant and purified antigen is no longer a ‘candidate’, but there would have been a reason for going to all the trouble to create it in a cell system.

Throughout: the authors should note that the lack of correlation for many of their targets is due to the fact that the purified antigen is maxing out the MFI signal whereas is seems IVTT targets never do this.

Author response:

We thank the reviewer for their comments and have included additional text highlighting the state-of-the-art with regards to the IVTT platform. The low signals described in this study and covered extensively in the literature is a known limitation of the system. The high throughput nature of the system is somewhat tempered by the high level of heterogeneity in protein production that is inherent with the system. The IVTT microarray approach is more of an ‘opt in’ where positives can be further validated using other approaches. This circumvents the heterogeneity which can result in detectable signal in one expression run vs no signal in another. This is fully covered in the following references and has been included in the introduction (Dopp et al Synthesis and systems biology 4, PMID: 31750411; Harini et al. Current Opinions in Biotechnology 2022, PMID: 35247659).

Although the system is scalable, in terms of volume of reaction mix/culture, this does not circumvent the known heterogeneity that is inherent within the system. This is an currently an area of active research. Large-scale systems do exist within the biotechnology industry, but these are not generally accessible to researchers due to space considerations and cost to run (Silverman et al Nat Rev Gen 2019, PMID: 31780816).

Refer to Line 46-57

Relevant controls are built into the array and are described under the heading “Protein microarray”.

Submitted filename: Responses to reviewers_PONE_D_21_37402_v2.docx

Click here for additional data file.

3 Aug 2022

Plasmodium falciparum serology: A comparison of two protein production methods for analysis of antibody responses by protein microarray

PONE-D-21-37402R1

Dear Dr. Tetteh,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Takafumi Tsuboi

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: (No Response)

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: (No Response)

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: (No Response)

Reviewer #2: (No Response)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

**********

19 Aug 2022

PONE-D-21-37402R1

Plasmodium falciparum serology: A comparison of two protein production methods for analysis of antibody responses by protein microarray

Dear Dr. Tetteh:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Prof. Takafumi Tsuboi

Academic Editor

PLOS ONE

Table 1

Description of P.falciparum antigens and their corresponding IVTT and purified protein targets.

Protein	Description	Full length (amino acids)	Protein target/expression system	Size (Start amino acid—End amino acid)
ACS5	Acyl CoA synthase	811	IVTT_1	811 (1−811)
			Pure_1	117 (294−410)
			Pure_2	160 (414−573)
			Pure_3	150 (578−727)
AMA1	Apical membrane antigen 1	622	IVTT_1	622 (1−622)
			Pure_1	450 (97−546)
EBA181	Erythrocyte binding antigen 181	1567	IVTT_1	754 (1−754)
			IVTT_2	752 (737−1488)
			Pure_1	585 (755−1339)
ETRAMP4	Early transcribed membrane antigen 4	136	IVTT_1	136 (1−136)
			Pure_1	25 (28−52)
			Pure_2	61 (76−136)
ETRAMP5	Early transcribed membrane antigen 5	181	IVTT_1	181 (1−181)
			Pure_1	86 (26−111)
			Pure_2	47 (135−181)
GAMA	GPI-anchored membrane antigen	738	IVTT_1	738 (1−738)
			Pure_1	99 (68−166)
HSP40	Heat shock protein 40 type II	402	IVTT_1	134 (80–213)
			IVTT_2	171 (213–401)
			Pure_1	83 (71–153)
			Pure_2	189 (214–402)
MSP1	Merozoite surface protein 1	1720	IVTT_1	870 (1–870)
			IVTT_2	868 (853–1720)
			Pure_1	45 (64–108)
			Pure_2	35 (54–63;109–133)
			Pure_3	116 (1605–1720)
MSP4	Merozoite surface protein 4	272	IVTT_1	162 (1–162)
			IVTT_2	161 (1–161)
			IVTT_3	110 (163–272)
			Pure_1	65 (43–107)
MSP5	Merozoite surface protein 5	272	IVTT_1	172 (1–172)
			IVTT_2	171 (1–171)
			Pure_1	61 (147–207)
MSP7	Merozoite surface protein 7	351	IVTT_1	351 (1–351)
			Pure_1	175 (177–351)