Adaptation to simulated microgravity in Streptococcus mutans.

Long-term space missions have shown an increased incidence of oral disease in astronauts' and as a result, are one of the top conditions predicted to impact future missions. Here we set out to evaluate the adaptive response of Streptococcus mutans (etiological agent of dental caries) to simulated microgravity. This organism has been well studied on earth and treatment strategies are more predictable. Despite this, we are unsure how the bacterium will respond to the environmental stressors in space. We used experimental evolution for 100-days in high aspect ratio vessels followed by whole genome resequencing to evaluate this adaptive response. Our data shows that planktonic S. mutans did evolve variants in three genes (pknB, SMU_399 and SMU_1307c) that can be uniquely attributed to simulated microgravity populations. In addition, collection of data at multiple time points showed mutations in three additional genes (SMU_399, ptsH and rex) that were detected earlier in simulated microgravity populations than in the normal gravity controls, many of which are consistent with other studies. Comparison of virulence-related phenotypes between biological replicates from simulated microgravity and control orientation cultures generally showed few changes in antibiotic susceptibility, while acid tolerance and adhesion varied significantly between biological replicates and decreased as compared to the ancestral populations. Most importantly, our data shows the importance of a parallel normal gravity control, sequencing at multiple time points and the use of biological replicates for appropriate analysis of adaptation in simulated microgravity.

Introduction

As NASA’s desire to explore and set up human habitation on planets like Mars increases, preventing and/or predicting potential threats to mission objectives are a main priority. Humans face many challenges in space[1,2]; one example being the many microbial challenges they encounter, potentially hindering their own level of performance, and the integrity of their spacecraft and habitat[3,4]. Microbes are not only ubiquitous on earth, but are also readily found on structures inhabited by humans in the spaceflight environment[3,5]. The crew’s normal flora harbors large quantities of microbes making them the most important source of bacterial contamination[6-9]. The human microbiome is essential for our survival as it helps break down food, protects us from pathogens and even primes our immune system. Under conditions of distress, the composition of the normal flora is often altered, ultimately leading to dysbiosis and disease[10-14]. This dysbiosis could be further enhanced by the immune dysregulation encountered during space travel[15-20]. Therefore, infection resulting from opportunistic pathogens is of concern during space flight[21].

Low-shear stress and reduced gravity can promote microbial dysbiosis and change bacterial physiology[4,22-24]. Experiments have shown that under simulated microgravity, microbes can evolve novel resistance[25], show enhanced growth[26], biofilm formation[27,28], extracellular polysaccharide production[29], increase production of secondary metabolites[30] while also showing alterations in pathogenic stress response and virulence[22,31-33]. To reduce incidence of infection in space, both the spacecraft and the crew undergo in-depth microbial screening prior to flight[34], but this mitigation strategy may not prevent novel evolutionary phenotypes arising within the crew’s microflora that will adapt due to their exposure to the novel environment of space.

Long-term space missions and increased exposure to simulated microgravity and radiation have shown an increased incidence of oral disease in astronauts[35]. As a result, The Space Medicine Exploration Medical Condition List indicates that diagnostic and treatment capabilities for basic dental procedures be made available[36]. In addition, the Integrated Mathematical Medical Model predicts that dental emergencies will be one of the top conditions impacting future mission objectives[37]. Approximately 20% of oral bacteria are streptococci, these organisms are responsible for both early establishment of dental plaque and dental decay[38,39]. The phenotypes of these organisms as they exist on earth are well studied and treatment strategies are more predictable, but these may not hold true during long-term space travel.

The human mouth is a very complex community made up of over 1000 different species and is the second most complex after the GI tract[40,41]. These communities reach cell densities as high as 1011 CFU mL−1[42] despite having to endure a constant change in environmental stressors including food intake, temperature, pH and salivary flow[43-45]. The oral microbiome plays a role in not only maintaining oral health, but also in maintaining systemic health as all surfaces of the oral cavity (teeth, gums, tongue etc) are inhabited by microbes which aids in preventing colonization by pathogens[46,47]. Despite this, dental decay remains one of the highest prevailing diseases in humans[48]. Of the oral microbiome residents, Streptococcus mutans has been actively studied for its cariogenic properties, as this organism not only causes dental decay, but resides as a member of normal human plaque[49].

S. mutans is a gram-positive facultative anaerobe (Firmicutes) that normally exists as a member of the mature dental biofilm community, but under certain conditions, can become the dominant species leading to dental caries[50,51]. The formation of dental caries is reliant on two factors, 1) an ecological shift that favors the growth of acid producing bacteria and 2) the presence of sucrose in the environment for both fermentation and production of glucans that facilitate attachment of the organism to the tooth, leading to formation of the plaque biofilm which is tolerant to low pH[52].

Two short-term studies assessed the impact of simulated microgravity on S. mutans. Orsini et al.[53], demonstrated using a High Aspect Ratio Vessel (HARV) for 48 h that S. mutans underwent both transcriptomic and metabolomic changes in carbohydrate metabolism and increased stress in the form of hydrogen peroxide susceptibility and noted variation between biological replicates. Orsini et al.[53] also showed an increase in cellular aggregates, indicative of an increase in cell-to-cell adhesion and/or biofilm formation. Subsequently, Cheng et al.[54] found that S. mutans displayed little changes in growth and hydrogen peroxide tolerance while exhibiting an increase in acid tolerance in response to simulated microgravity. It is hard to determine the value of these short-term studies (48 h) as a typical mission can last 4–6 months at a time. It is highly likely that these short-term studies only represent potential physiological acclimation, as opposed to evolutionary adaptation[55].

Currently, we have limited knowledge of microbial adaptation in response to long duration space flight. The common human resident: Streptococcus mutans, will be taken by every astronaut into space[49]. Therefore, studies such as the present one are essential for predicting which organisms have the potential to evolve into strains with increased virulence and pose a greater risk to human health once in space.

Experimental evolution followed by whole genome resequencing (EERseq) experiments are commonly used for evaluating the genomic changes associated with selection regimes[56-61]. Tirumalai et al.[62] adapted E. coli in simulated microgravity for 1000 generations (~50 days) using HARVs and showed it acquired genomic changes in genes involved in outer membrane protein folding, ion transport and in the stress response. In a similar experiment, they also evaluated the consequences of periodic exposure to antibiotic therapy[63]. To date, these are the only long-term microbial adaptation studies in simulated microgravity in the literature[62,63]. Unfortunately, likely due to the necessity of specialty equipment, adaptation experiments to simulated microgravity are often under-powered with few biological replicates (populations evolved in parallel to capture random variation) and often lacking normal gravity populations for comparison. The absence of a normal gravity control means that Tirumaliai et al.[62,63] could not legitimately claim that the variants that arose within their simulated microgravity treatment were not just adaptations to some other aspect of their environment, such as the medium.

In the present study, we used HARVs to evaluate the adaptive response of four biological replicates of S. mutans to simulated microgravity[64] over 100-days (~1400 generations) to better understand the consequences of long-term space travel on organisms that reside as normal residents of the host microbiome. The adaptive response was evaluated by performing whole-genome resequencing every three weeks and phenotypes correlated with virulence were assessed after 100-days of adaptation. All adapted populations were compared to the ancestral population as well as to their normal gravity counterparts which differ only by the axis of rotation of the vessel itself[4,65]. Understanding the long-term evolution of the human microbiome in outer space will therefore be an important step in further understanding the effects of space travel on humans and their resident microbes.

Methods

Culture strains

Streptococcus mutans Clarke strain NCTC 10449 was purchased from the ATCC [https://www.atcc.org/products/25175]. All standard growth experiments were conducted in Brain Heart Infusion (BHI) broth (or agar) at 37 °C with 5% CO2 unless otherwise noted.

Preparation of HARV vessels

High Aspect Rotating Vessels (HARVs) were purchased from Synthecon Inc., Houston, TX and used to culture S. mutans under simulated microgravity. Prior to inoculation, the HARVs were cleaned using a mild dish soap and water and rinsed in distilled water twice. All the components were then soaked in a 25% bleach solution for 15 min, rinsed extensively in tap water and then rinsed in distilled water. The HARVs were then loosely reassembled (as per the manufacturer), wrapped in aluminum foil and autoclaved at 121 °C for 20 min. HARVs were then left to cool for 2 h. After they cooled, HARVs were loaded with 5 mL of BHI broth and placed onto the rotator backplate at 25 rpm for 24 h, at 37 °C and 5% CO2 to ensure that the HARVs were sterile. This procedure was repeated anytime that contamination was detected over the course of the 100-day evolution experiment.

Experimental evolution (EE) of S. mutans over 100-days of LSMMG exposure

The physical and mechanical unloading by simulated microgravity in ground-based systems have been conducted in many studies, characterizing how simulated microgravity impacts various organisms and biological systems[4,6,15,22-27,29,53,54,62,66-73]. An overview of the experimental methods is depicted in Fig. 1. S. mutans Clarke strain NCTC 10449 was used to inoculate 3 mL of fresh BHI broth and incubated overnight with shaking at 250 rpm. This stock was then used to streak a BHI agar plate and left to incubate overnight. A single colony was then used to make a glycerol stock deemed the ancestral population. Initial growth curves of the ancestral population in the HARVs showed saturation after ~24 h and the generation time was determined to be ~14 per 24 h (one generation per ~105 min). To begin the EE protocol the ancestral stock was used to start another overnight culture. 100 μL of this overnight culture was then used to inoculate 100 mL of fresh BHI broth, 10 mL of the sub-culture was then loaded into a 10 mL sterile syringe and screwed into one of the two openings on the HARV (inlet), a second empty syringe was placed on the second opening (outlet). 10 mL was then pushed into the inlet while the syringe on the outlet collected the media from the HARV. This was repeated for all 8 HARVs. Four of the HARVs were then incubated on the vertical axis of rotation perpendicular to gravity and deemed normal gravity to serve as the controls and the other four were incubated on the horizontal axis to simulate simulated microgravity. All eight HARVs were incubated at 37 °C overnight with 5% CO2 at 25 rpm. After 24 h of growth (~14 generations, Supplementary Fig. 3), the HARVs were then subcultured by adding 10 mL of fresh BHI into the inlet port which pushed the culture from the HARV into an empty syringe attached to the outlet port and returned to the incubator for a new 24 h cycle. This was carried out daily for 100-days (~1400 generations). During the EE study, the HARVs were checked daily for contamination by first, measuring the O.D.600, values greater than 1 were often indicative of contamination, second, we performed a simple stain with crystal violet on an aliquot of the culture and observed it under a compound light microscope for general shape and arrangement and anything that was uncharacteristic of S. mutans. Then twice a week we made glycerol stocks and plated serial dilutions onto both BHI agar and Mitis Salivarius Bacitracin agar (MSB) which is both selective and differential for S. mutans. These plates were used to validate the integrity of the glycerol stock which would be used in case of future contamination. Every milestone time point (21-, 42-, 63- and 100-days), the remainder of the culture were pelleted and stored at −80 °C for DNA sequencing. If contamination was detected we would sterilize the HARVs as previously described, inoculate with fresh media for 24 h and then inoculate with the most recently validated glycerol stock.

Fig. 1

Experimental methods.

Schematic representation of the experimental evolution workflow used to adapt Streptococcus mutans to simulated microgravity.

Etest® analysis for measuring antibiotic susceptibility

All eight 100-day populations, in addition to the ancestral, were used to swab BHI plates in triplicate. On each of those plates, one Etest® strip (bioMériux) was placed in the center of the plate and left to incubate for 48 h to obtain confluent lawns. Antibiotic susceptibility was then measured by determining the minimum concentration on the strip at which growth was observed. This value was then compared for each of the eight evolved populations against the values obtained for the ancestral population by performing a one-way ANOVA in GraphPad Prism using the multiple comparisons function with the ancestral. In total, six antibiotics were tested including: amoxicillin, penicillin, clindamycin, erythromycin, methicillin, and vancomycin.

Acid tolerance assays

Acid resistance is one of the main virulence traits in S. mutans, we therefore assessed our evolved populations for changes in their ability to survive at low pH. Acid tolerance assays were performed as described elsewhere[74]. In short, after removal of the cultures from the HARVs, a sample of each population was taken and diluted to an O.D.600 of 0.3. These cultures were then centrifuged to pellet the cells and then washed using 0.2 M glycine pH 6.8. Samples were again pelleted and exposed to 0.2 M glycine pH 2.8 for 0, 20, 30 and 45 min. At the end of the incubation time, samples were pelleted and resuspended in 0.2 M glycine pH 6.8. Serial dilutions of each time point were then plated on BHI agar and CFUs were counted after 24 h of incubation and compared to CFU counts of the ancestral population. Each dilution was plated in triplicate to acquire accurate CFU counts.

Adhesion assays

The ability to adhere to the tooth pellicle is another main virulence factor required by S. mutans to establish itself in a biofilm. Therefore, at 100-days, all eight subcultured populations were assessed for changes in their adhesion abilities. Cultures were first diluted to an O.D.600 of 0.05 and 5 μL were used to inoculate 95 μL of fresh BHI broth supplemented with 0.1% sucrose to measure sucrose dependent adhesion or 0.1% glucose to measure sucrose-independent adhesion in a 96-well plate, in triplicate. The plates were then incubated without shaking at 37 °C for 24 and 48 h with 5% CO2. After 24 h, the media was removed, and the plate was washed with distilled water. 200 μL of a 1% crystal violet solution was then added to the washed plate and left to incubate for 2 h. Plates were then washed with water three times and left to dry overnight. The following day, 200 μL acetic acid was used to release the crystal violet from adhered cells and then was read at 595 nm. Each population was plated in three separate wells to acquire three independent measurements for each population.

Genome sequencing

Chromosomal DNA was extracted from the pellets originating from the 21-, 42–63- and 100-day samples for all eight evolved populations and the ancestral, using the E.Z.N.A. bacterial DNA extraction kit from Omega Biotek® as per the manufacturer’s protocol. Eluted DNA was then quantified using the QuantiFluor® dsDNA system (Promega). 300 ng of purified genomic DNA was used to prepare DNA libraries using the Nextera XT DNA library prep kit (Illumina). Samples were then sequenced on an Illumina miSEQ sequencing platform with depth of coverage ranging from ~100X to ~200X, with most exceeding 150X coverage. Sequence alignment and variant calling from the samples was achieved by use of the breseq 0.30.0 pipeline[75]. The breseq pipeline uses three types of evidence to predict mutations, read alignments (RA), missing coverage (MC), and new junctions (JC), and any reads that indicate a difference between the sample and the reference genome that cannot be resolved to describe precise genetic changes are listed as “unassigned.” These unassigned reads are not described nor interpreted here.

Statistics

All phenotypic data was plotted in GraphPad Prism ® Version 9.2.0. Statistics for pairwise comparisons between ancestral populations and each individual treatment population were calculated using an unpaired T-test. Significance is defined by the two-tailed p-value using an *. * ≤ 0.05, ** ≤ 0.01, *** ≤ 0.001 and **** ≤ 0.0001. Linear regression was used to compare the number of accumulated mutations for each normal gravity population and each simulated microgravity population. General linear analysis in SPSS was used to compare the normal gravity and simulated microgravity populations. Finally, chi-squared analysis using SPSS was used to compare the genomic variants to determine the effect of environment on selection and a simple linear regression was performed in GraphPad Prism with a 95% confidence interval to determine the significance in the slope for reporting of accumulated mutations in both normal gravity and simulated microgravity.

Table 1

Frequency of mutations detected earlier in simulated microgravity populations.

Gene	Day	NG1	NG2	NG3	NG4	MG1	MG2	MG3	MG4	Annotation
SMU_399	21					0.10				S112Y
C3-GDP							0.06	0.26		C 302–313/759
									0.14	C 24–36/759
						0.10				E58*
	42	0.29								C276–514/759
		0.09					0.58	0.88		C 302–313/759
						0.47				S112Y
						0.09				E58*
							0.10			E221*
									0.58	C 538/759
	63	0.30								C 276–514/759
		0.08					0.85	0.63	1.00	C 302–313/759
						1.00				S112Y
				0.11						Q130*
								0.09		C 664/759
								0.23		C 29/759
	100	1.00		1.00			1.00			C 7/759
			1.00		0.26		0.17	0.66	1.00	C 302–313/759
						0.86				S112Y
						0.10				E58*
							0.09			E221*
ptsH	42					0.19	0.50			G54A
	63		0.33		0.21	0.43	0.72	0.34	1.00	G54V
	100				0.06	0.06			0.93	G54V
rex	42					0.13				Q202*
						0.06				T155K/G65S/D52N/
										R51L/S46La
							0.22			Y66C
								0.83		Q193P
									0.66	A33E
	63	0.25								R51H
			0.07							R14H
				0.26						T155R
				0.61						T48I
					0.05					G60S
					0.10					A47V
						0.08				G57S
						0.37				A33E
							0.13			Y66C
								0.53		Q193P
								0.05		A115E
	100				0.20					R51L
							0.58			Y55D

*Indicates premature stop codons.

C indicates in the coding region.

aMG1 acquired 5 different mutations all at an f ~0.06.

Table 2

Frequency of mutations detected earlier in normal gravity populations.

Gene	Day	NG1	NG2	NG3	NG4	MG1	MG2	MG3	MG4	Annotation
pknB	42			0.05						G174C
				0.13						R45C
					0.11					Y475*
	63	0.34								Y561*
		0.11								D83H
				0.28	0.26					R45C
							0.08			D78E
									0.86	R258C
	100				0.13					F58S
					0.35					R45C
						0.16				I2S
									0.09	R258C
									0.60	G19A
DUF1003 domain containing protein	21	0.07	0.14							L168I
	42	0.07								R181C
		0.10								S231R
									1	A157V
	100	0.65								E195G
		0.17								R199C
			0.93							R181C
			0.94							S231R
				0.28						Q100*
				0.54						R140*
					0.24			0.52		D96E
						0.12				K102Q
						0.47				T235R
							0.11			L168I
								0.11		A166S
								0.13		A166D
									0.92	A157V
DQM59_RS01880	100			0.90						E244*
Cof‑type HAD‑IIB family hydrolase
					0.26			0.53		A242E
					0.15					G54A
				0.90						E244*

*Indicates a STOP codon.

Table 3

Frequency of variants unique to normal gravity.

Day	NG1	NG2	NG3	NG4	Annotation	Gene
21			0.11	0.24	F52V	DQM59_RS10195 ← (sprV)
			0.27	0.18	H21N	DQM59_RS10195 ← (sprV)
42			0.38	0.14	Q31*	DQM59_RS04330 → (ridA)
	0.08				R34S	DQM59_RS10195 ← (sprV)
		0.05			E72K	DQM59_RS10195 ← (sprV)
		0.24			K15N	DQM59_RS10195 ← (sprV)
			0.08	0.17	F52V	DQM59_RS10195 ← (sprV)
			0.12	0.13	H21N	DQM59_RS10195 ← (sprV)
		0.10			T104A	vicK
	0.74	0.15			C 2051–2131/4689	spaP ←
				0.08	S721L	spaP ←
	0.35				K373N	DQM59_RS07770 ←(murD)
63			0.28	0.46	Q31*	DQM59_RS04330 → (ridA)
		0.71			F219L	DQM59_RS06715 ←
	0.22				K373N	DQM59_RS07770 ←(murD)
	0.25				R34S	DQM59_RS10195 ← (sprV)
	0.08				H21Y	DQM59_RS10195 ← (sprV)
		0.54			K15N	DQM59_RS10195 ← (sprV)
			0.58		F52L	DQM59_RS10195 ← (sprV)
			0.10		F52V	DQM59_RS10195 ← (sprV)
				0.36	H21N	DQM59_RS10195 ← (sprV)
					E55D	DQM59_RS10195 ← (sprV)
	0.22				K7Q	DQM59_RS02455 →
	0.59	0.05			C 2051–2131/4689	spaP ←
			0.08		S721L	spaP ←
				0.09	N917N	spaP ←
		0.21		0.05	R205C	DQM59_RS03385 →
100				0.55	Q31*	DQM59_RS04330 → (ridA)
				0.33	I(+280/−96)	DQM59_RS00680 → / → DQM59_RS00685
			0.53		F98V	DQM59_RS03205 →
		1.00			I(−236/−20)	acnA ← / → DQM59_RS07205
		0.62			C 464–493/495	DQM59_RS01355 →
		0.93			P169L	DQM59_RS03185 →(trkA)
				0.18	S191*	DQM59_RS07640 →(lytS)
				0.15	Q229*	DQM59_RS07640 →(lytS)
	0.11				K15N	DQM59_RS10195 ← (sprV)
			0.06		F52V	DQM59_RS10195 ← (sprV)
			0.60		V81F	vicK
			0.38		I407F	vicK
				0.26	A237D	vicK
				0.20	S721L	spaP ←

*Indicates premature stop codons.

C indicates in the coding region.

I indicates intergenic region.

Table 4

Frequency of variants unique to simulated microgravity.

Day	MG1	MG2	MG3	MG4	Annotation	Description
42	0.06				G88A	rpoC
63		0.20			G123C	ccpA →
	0.12				Y636*	DQM59_RS06275 ←
		0.12			A78V	DQM59_RS04610 →(eno)
	1.00				G88A	rpoC
		0.06			E105G	rpoC
		0.12			E55D	DQM59_RS10195 ← (sprV)
			0.11		L22I	ylqF ←
100				0.14	intergenic (+181/+120)	asnS → / ← DQM59_RS04320
		0.20			G77R	DQM59_RS00400 →
	0.84				intergenic (−381/−244)	DQM59_RS00535 ← / → DQM59_RS00540
			0.15		V96G	DQM59_RS03190 →(trkB)
			0.15		D80A	DQM59_RS03980 →(lepA)
			0.15		L81F (	DQM59_RS03980 →(lepA)
		0.13			M267L	DQM59_RS05315 ←(vex3)
			0.32		pseudogene (217/304 nt)	DQM59_RS08775 ←
	0.88				G88A	rpoC
		0.65			H21N	DQM59_RS10195 ← (sprV)
			0.37		Y88C	DQM59_RS10195 ← (sprV)
			0.41		A237D	vicK
		0.15			W443*	vicK
			0.19		coding (1588–1701/4689)	spaP ←
		0.11			E251*	gorA ←

*Indicates premature stop codons.

C indicates in the coding region.

I indicates intergenic region.