规律间隔成簇短回文重复序列（clustered regularly interspaced short palindromic repeats，CRISPR）是原核生物中的一种特殊的重复序列。它由富含腺嘌呤和胸腺嘧啶的前导序列（leader）和由独特间隔序列（spacer）分隔的短重复序列组成。CRISPR与CRISPR关联蛋白（CRISPR-associated proteins，Cas proteins）组成的CRISPR-Cas系统是一种在原核生物中天然存在的适应性免疫系统，可以保护细胞免受病毒和质粒等可移动遗传元件（mobile genetic element，MGE）的入侵。

相较于其它类型的免疫系统而言，CRISPR-Cas系统在原核生物中的分布是十分不均匀的。CRISPR-Cas系统存在于约85%的古细菌基因组中，但是仅有约40%的细菌中含有CRISPR-Cas系统。以往的研究表明CRISPR-Cas系统在嗜热菌和超嗜热菌中的普遍存在，并且CRISPR-Cas丰度与生长温度之间存在正相关关系。本研究观察到细菌的CRISPR-Cas系统丰度在最适生长温度45℃时突然增加。通过系统发育比较分析进一步证实，这种丰度与最适生长温度的相关关系仅在45℃左右的范围内出现。以往的研究表明，细胞捕食者（例如原生动物、线虫和粘细菌等）的多样性在这一温度线急剧下降。细菌面临的被捕食风险在45℃左右显著降低，在60℃以上则几乎消失。为此本文提出假说：在最适生长温度较高的物种中，病毒裂解是细菌死亡的主要因素，所以抗病毒的免疫具有更高的优先级。

本研究选择了CRISPR重复序列的长度、RNA二级结构最小自由能、数量和保守度这四个特征，分析了它们与生长温度之间的关系。本研究发现CRISPR重复序列RNA二级结构最小自由能（取值范围负数至0）与最适生长温度正相关，即CRISPR重复序列RNA二级结构的折叠程度（稳定性）与最适生长温度负相关。这一结论与以往对于RNA结构与温度关系的研究相符。细菌和古细菌的CRISPR重复序列长度与最适生长温度之间没有显著的相关性。虽然温度一直以来都被认为是自发突变的主要决定因素，但是本研究并没有观察到最适温度高的物种中CRISPR重复序列保守度降低。在PGLS回归分析中，本研究发现古细菌CRISPR重复序列数量与最适生长温度存在显著的负相关。本研究推测这可能是由于与细菌相比，古细菌的温度分布的上限更高，随着古细菌最适生长温度升高进入极端高温后，环境中病毒的多样性下降，随之CRISPR的重复序列（间隔序列）数量也出现了下降。

水平基因转移（horizontal gene transfer，HGT）是原核生物基因组扩增的主要驱动力。病毒和质粒等可移动遗传元件是水平基因转移的主要手段，而CRISPR-Cas系统则是针对外源可移动遗传元件适应性免疫工具。以往的研究也指出CRISPR免疫系统靶向外来DNA会在一定程度上限制水平基因转移。本研究通过PGLS回归分析发现，细菌的相关病毒和质粒存在与否及数量均显著与基因组大小正相关。古细菌中质粒与基因组大小显著正相关，但病毒与基因组大小没有显著的相关性。本研究通过PGLS回归分析和系统发育成对比较发现，原核生物基因组的CRISPR有无均与基因组大小不相关。以涉及水平基因转移的基因数量为指标，细菌的水平基因转移程度与相关病毒和质粒存在与否及数量呈显著正相关但与CRISPR有无不相关。本研究推测CRISPR-Cas系统抑制水平基因转移程度较弱，总体上可能并不会对水平基因转移的倾向性产生显著的影响。CRISPR-Cas系统在种群水平对水平基因转移产生的影响可能并不足以在物种进化尺度上造成差异。

综上所述，本研究分析了CRISPR-Cas系统丰度在原核生物中的温度分布，发现细菌的CRISPR-Cas系统丰度在45℃左右出现激增，并根据被捕食风险的变化提出了假说。本研究还发现CRISPR重复序列RNA二级结构稳定性与最适生长温度负相关、CRISPR存在与否与原核生物基因组大小不相关等现象。本研究的发现和假说为微生物生态学、基因组进化以及CRISPR-Cas系统在原核生物中适应性作用的研究提供了一些新思路。

CRISPR (clustered regularly interspaced short palindromic repeats) is a special type of repetitive DNA sequences found in prokaryotes, consisting of an AT-rich leader sequence followed by short repeats that are separated by unique spacers. The CRISPR-Cas system, including CRISPRs and CRISPR-associated proteins (Cas proteins), is an adaptive immune system that occurs naturally in prokaryotes. It protects host’s cell from invasion by mobile genetic elements (MGEs) such as viruses and plasmids.

Compared to other immune systems, the distribution of the CRISPR-Cas system in prokaryotes is very heterogeneous. Around 85% of archaeal genomes contain the CRISPR-Cas system, while only about 40% of bacteria contain it. Prior research has demonstrated that the CRISPR-Cas system is prevalent in thermophiles and hyperthermophiles, and there is a positive correlation between the abundance of CRISPR-Cas system and growth temperature. New findings of this study show precipitous abrupt increase of bacterial CRISPR-Cas abundance at around 45℃. We have confirmed through phylogenetic comparative analyses that this abundance only correlates with growth temperature in the range around 45℃. Previous studies have shown that the diversities of cellular predators (such as protozoa, nematodes and myxobacteria) decline sharply within this temperature range. The risk of grazing faced by bacteria decreases significantly at temperatures around 45℃ and is almost completely eliminated above 60℃. We propose that viral lysis would become the predominant factor in bacterial mortality, and antiviral immunity would become a higher priority at higher temperatures.

This study selected four features, including length, minimum free energy of RNA secondary structure, count, and conservation of CRISPR direct repeats, to investigate their relationship with growth temperature. We found that the minimum free energy of RNA secondary structure of CRISPR repeat sequences (from negative to 0) was positively correlated with the optimal growth temperature, i.e., the level of folding (stability) of RNA secondary structure of CRISPR repeat sequences was negatively correlated with the optimal growth temperature. This finding is consistent with previous studies on the relationship between RNA structure and temperature. We did not observe a significant correlation between CRISPR direct repeat length and optimal growth temperature in bacteria and archaea. Although temperature has long been recognized as a major determinant of spontaneous mutation, we did not observe reduced CRISPR direct repeat conservation in species with high optimal temperatures. In PGLS regression analysis, we found a significant negative correlation between the archaeal CRISPR direct repeats count and the optimal growth temperature. We postulate that this may be due to the higher upper limit of the temperature distribution in archaea compared to bacteria. As the optimal growth temperatures of archaea increase into extreme high temperatures, there is a decrease in the diversity of viruses in the environment, and consequently a decrease in the counts of CRISPR direct repeats (spacers).

Horizontal gene transfer is a major driving force of genome expansion in prokaryotes. Mobile genetic elements, such as viruses and plasmids, are primary facilitators of horizontal gene transfer, while the CRISPR-Cas system represents an immune mechanism that targets exogenous mobile genetic elements. Previous studies have also pointed out that the CRISPR immune system targeting foreign DNA would limit horizontal gene transfer to some extent. Our study indicates that the presence and number of associated viruses and plasmids in bacteria were significantly and positively correlated with genome size by PGLS regression analysis. Plasmids were significantly positively correlated with genome size in archaea, but viruses were not. We found that neither the presence nor absence of CRISPR in prokaryotic genomes was correlated with genome size by PGLS regression analysis and phylogenetic pairwise comparisons. Using the number of genes involved in horizontal gene transfer event as an indicator, the horizontal gene transfer in bacteria was significantly positively correlated with the presence and number of associated viruses and plasmids but not with the presence or absence of CRISPR. We speculate that the CRISPR-Cas system's ability to inhibit horizontal gene transfer may not be strong enough to fundamentally impact the propensity for horizontal gene transfer. CRISPR-Cas system’s influence of HGT on population scale may not have enough leverage to make a difference on the timescale of evolution.

In summary, this study analyzed the temperature distribution of CRISPR-Cas system abundance in prokaryotes, and found that the abundance of CRISPR-Cas system in bacteria showed a surge around 45°C, and put forward a hypothesis based on the change of predation risk. We also found that the stability of RNA secondary structure of CRISPR repeat sequences is negatively correlated with the optimal growth temperature, and the presence or absence of CRISPR is not correlated with the genome size of prokaryotes. The new findings and hypothesis of this study provide new insights into microbial ecology, genome evolution, and the adaptive role of the CRISPR-Cas system in prokaryotes.