Response of pneumococcus to changes in temperature and oxygen varies by serotype, lineage, and site of isolation

Background Pneumococcus, a commensal bacterium of the nasopharynx, has the potential to cause severe disease. Invasion of different body sites results in exposure to a range of environmental conditions, including different temperature and oxygen levels. The response to these variations could influence virulence and transmissibility. Methods We evaluated the effect of temperature and oxygen on the growth of 256 pneumococcal isolates representing 53 serotypes, recovered from healthy carriers and from invasive disease (IPD), conjunctivitis, and pneumonia patients. Strains were grown at a range of physiologically-relevant temperatures anaerobically or in ambient air with and without catalase and were monitored by reading the optical density. Results There was considerable variability between strains in response to changes in temperature and oxygen. Some of this variability was associated with serotype. Generally, carriage and IPD isolates grew to the maximal density at the temperature of the nasopharynx (∼33C), but IPD isolates began growing earliest at internal body temperature. Growth curve characteristics correlated with epidemiological estimates of serotype-specific carriage prevalence and invasiveness. Discussion Environmental variability has differential effects on the growth of carriage and disease isolates and of different pneumococcal serotypes, which could influence virulence and transmissibility.

monitored by reading the optical density.

49
Results: There was considerable variability between strains in response to changes in temperature 50 and oxygen. Some of this variability was associated with serotype. Generally, carriage and IPD 51 isolates grew to the maximal density at the temperature of the nasopharynx (~33C), but IPD 52 isolates began growing earliest at internal body temperature. Growth curve characteristics 53 correlated with epidemiological estimates of serotype-specific carriage prevalence and 54 invasiveness.  Streptococcus pneumoniae (pneumococcus) is a commensal bacterium of the human nasopharynx. 62 The nasopharynx is considered to be the reservoir and source of pneumococcal transmission 63 between individuals [1]. Pneumococcus is an opportunistic pathogen that causes diseases ranging 64 from mild but common illnesses like sinusitis and otitis media to serious illnesses such as 65 pneumonia, sepsis and meningitis [2].

67
In different anatomical sites within the human host, pneumococci are exposed to variable 68 temperature and oxygen levels. In the nasopharynx, considered its main niche, the average 69 temperature is around 33°C, with some differences between children and adults [3-6]. The core 70 body temperature, which would be encountered during invasion into tissues, is 37°C. The 71 temperature in the lungs is constantly changing based on the temperature of inhaled air but is 72 generally lower than 37°C [7]. During infection by pneumococci or during viral co-infection (such 73 as influenza or RSV), both external and internal temperature increases [8][9][10][11][12]. Oxygen levels also 74 vary within the host. In the nasopharynx, bacteria on top of the mucus layer are exposed to almost 75 ambient air (20% O2). Pneumococci in biofilms in the nasopharynx encounter lower levels of 76 oxygen [13]. Entering the lower respiratory tract or the middle ear, pneumococci are exposed to 77 micro-aerophilic conditions and to almost anaerobic conditions when present in blood and the 78 cerebrospinal fluid (CSF) [14][15][16]. Likewise, mucus production during infection (i.e., due to 79 influenza or RSV) can block the air passage and form micro-aerophilic (around 5% O2) or even 80 anaerobic microenvironments [15,16]. To understand how pneumococci successfully colonize the host and transition from 83 nasopharyngeal colonization to severe disease, it is important to understand how different strains 84 6 respond to variations in the microenvironment. Pneumococcus is a diverse pathogen, with >90 85 serotypes (defined by the capsule polysaccharide) and has tremendous genetic variation that results 86 from recombination. This genetic variability influences the biology and epidemiology of the strains 87 [17]. Serotypes vary in their prevalence among healthy carriers and in the likelihood that they will  The aim of this study was to investigate how environmental variability in temperature and oxygen 96 influences the growth of different strains and serotypes of pneumococci. Using a diverse set of 97 clinical and nasopharyngeal isolates, as well as capsule-switch and capsule-knockout variants 98 generated in the lab, we quantified how the growth of pneumococcus in vitro varies under a range 99 of physiologically-relevant temperatures that represent colonization and invasive infections as well 100 in aerobic and anaerobic conditions. We evaluated the relationship between these in vitro growth 101 patterns and relevant serotype-specific epidemiological and biological variables. Additional capsule-knockout strains were generated by replacement of the capsule biosynthesis 114 locus with the Sweet Janus cassette [24].    Each growth curve was blanked by subtracting the OD600 reading at t=0 for that well. In instances 137 where the t=30 minutes measurement was lower than the t=0 measurement due to measurement 138 error at the first time point, the OD600 at t=30m was subtracted instead. For each growth curve, 139 we estimated the length of the lag phase and the maximum growth rate using the groFit package  Certain interactions among the fixed effects were also evaluated to test specific hypotheses (site 147 of isolate*oxygen; oxygen*temperature; serotype*oxygen). The significance of these interactions 148 at different levels was evaluated using the interactionMeans function in the phia package in R [27]. 149 Correlations of the effect of oxygen (ratio of maximum OD at a particular temperature for a 150 particular isolate grown anaerobically or aerobically with catalase) with previously described 151 epidemiological characteristics (invasiveness, carriage prevalence) were assessed using 152 Spearman's correlations [28]. Associations with the presence of specific capsule components were 153 assessed using Wilcoxon rank sum tests in R.

155
Principal components analysis and linkage to serotype-specific characteristics 156 We sought to evaluate the link between growth curve characteristics and previously published 157 serotype-specific characteristics (capsule structure, disease severity, invasiveness, pre-vaccine 158 carriage prevalence) [29-31]. Across all growth curves, isolates, conditions, and time points, there 159 was a large volume of data with many possible variables that could be generated for correlation 160 analyses (e.g., maximum OD600 achieved, OD600 at time point T). To avoid overfitting the data,   temperatures but grew to a significantly higher density in aerobic+catalase at low temperatures.

206
Conjunctivitis and pneumonia isolates also grew similarly in anaerobic and aerobic+catalase 207 conditions, although fewer isolates were tested (Figure 2).

209
When grown in ambient air without catalase, pneumococcus typically produces toxic levels of 210 hydrogen peroxide. As expected, most IPD isolates hardly grew in ambient air due to this toxicity.

211
Unexpectedly, some IPD isolates carriage isolates grew moderately under these conditions (e.g.,

212
IPD isolates of serotypes 9N, 12F, 18A,23A, 33F IPD isolate; Figure 5). Moreover, carriage 213 isolates grew moderately well in ambient air without catalase, especially at higher temperatures 214 (35-39°C). (Figure 2C, 3). The clinical isolates represented diverse genetic backgrounds. We attempted to evaluate the growth 226 characteristics of capsule-knockout strains as well as several capsule-switch variants. While the 227 results were ambiguous, they suggested that the effect of capsule production on growth was more 228 pronounced during anaerobic growth compared with aerobic growth with catalase 229 (Supplementary Figures 2-5). Finally, we performed exploratory analyses to evaluate whether the characteristics of the growth 233 curves were related to serotype-specific epidemiological characteristics and biochemical 234 characteristics of the capsule. There was an association between the second principal component 235 (PC2, which explained 11% of the variability in the data) and pre-vaccine carriage prevalence 236 (p<0.001) and invasiveness (p=0.05) (Figure 6). We therefore evaluated which components of the 237 growth curves contributed most to PC2. PC2 was largely influenced by density at ~5 hours among 238 the carriage isolates at 33C and ~2-3 hours among the IPD isolates at 38°C. These time points 239 corresponded to periods when many of the isolates were in early log-phase growth but some were 240 still in stationary phase. PC2 was also influenced by density at later time points, which largely 241 (inversely) reflects density at early time points.

243
There was no association between PC1-6 and serotype-specific case fatality ratio, polysaccharide  isolates responded differently to environmental variations. While they also reached maximum 266 density at low temperatures, the IPD isolates had the shortest lag phase at 37°C (i.e., core body 267 temperature). Additionally, IPD isolates hardly grew in ambient air without catalase, while 268 carriage isolates grew moderately well, especially at higher temperatures (Figure 1-4). This study had certain limitations. For the growth curves, we used BHI broth which is an artificial 309 growth medium that differs in nutrient composition from the host. We evaluated several minimal 310 media but found that growth was generally poor, making comparisons between strains difficult.

311
While we tested a large number of strains representing many serotypes, some serotypes were only 312 represented by a single isolate (i.e., 11B, 12F, 13). This could limit the generalizability of serotype-313 specific findings in these instances, making it difficult to make inferences about whether variability 314 was due to serotype, site of isolation, or lineage effects. The strains used in this study were largely with the capsule-knockout strains and capsule-switch variants suggests that the capsule itself could 318 influence these phenotypes. We did not perform any gene expression studies which could be highly 319 influenced by environmental conditions [39]. Further work could explore the genetic basis (both 320 capsular and non-capsular factors) for the differences in growth phenotypes between strains.

322
In conclusion, we demonstrate that the growth characteristics of pneumococcus are influenced by