To better understand aging, we studied intestinal bacterial accumulation in C. elegans differing in the bacterial species that they ingest, as well as their genotype and maturation. Here, we provide evidence that the extent of intestinal bacterial accumulation early in adulthood, which is controlled by gut immunity that decreases with age, is strongly and inversely correlated with longevity.
Bacteria are the source of nutrition for C. elegans, but ultimately as the worms age, viable bacteria accumulate in the intestine . Worms grown on the soil bacterium Bacillus subtilis have a longer lifespan compared to those grown on E. coli OP50 or many other tested bacterial species . However, worms that are grown on B. subtilis spores produce fewer eggs and are smaller and thinner than those fed on vegetative cells of B. subtilis or E. coli OP50 . This observation indicates that growth on spores compared to vegetative (metabolically active) bacterial cells limits nutrient availability. Thus, vegetative bacteria represent two competing elements to C. elegans: a nutrient that fosters development and fecundity, and a toxic component that may reduce lifespan . Worm defenses, including the pharyngeal grinder and intestinal immunity, act to mitigate the latter phenomenon.
The nematode responds to bacteria with conserved innate immune responses, however, aging is accompanied by a decline of immune functions [18, 19]. This may represent a general evolutionary process, since after reproductive age individuals compete with their own progeny for available nutrients. Although the functionality of the C. elegans immune system during aging has been extensively examined [38, 63], we now have simultaneously examined longevity and control of bacterial proliferation across worm genotype, age, and bacterial strain differences. We confirm that viable bacteria accumulate in the C. elegans intestine as they age , and now show that both bacterial strain type and worm genotype related to gut immunity affect intestinal bacterial accumulation, which might play a significant role in lifespan determination, since we found that lifespan and bacterial load are inversely correlated. Previous studies had quantified bacterial proliferation by CFU enumeration only in N2 worms . More recent studies showed substantially fewer bacteria in the gut of certain long-lived C. elegans mutants; however, these observations were by semi-quantitative microscopy only . By quantitatively characterizing the kinetics of bacterial proliferation in the C. elegans intestine, in wild type and mutant worms, we establish a basis to better dissect the interplay of bacteria, host genotypes, and age.
One of the aims in this study was to characterize the kinetics of intestinal bacterial colonization. Salmonella is a pathogen of C. elegans that permits examining this question since it kills worms relatively slowly, rather than in a rapid manner. However, other than consistently higher numbers, there were few cases in which Salmonella and E. coli results differed greatly. These differ from previous data that reported significant differences in the lifespan of C. elegans when grown on Salmonella compared to E. coli . The discrepancy might be explained in part by differences in methodology, since in this work we grew the worms on lawns of Salmonella rather than exposing them as L4's. However, E.coli also is pathogenic to C. elegans [15, 31, 64], and many C. elegans antimicrobial genes are induced, some even more strongly (lys-1 and spp-1) than in the presence of other pathogens . As such, E. coli is just one other bacterial species to which C. elegans can sense and respond.
In our experimental system, we found significant differences in bacterial accumulation at day 2 of adult life, and that variation in the intestinal bacterial loads among the immunodeficient mutants correlated with lifespan differences. Why were differences in bacterial proliferation significant at day 2? One explanation is that since C. elegans produces nearly all of its progeny within the first 2 days of its adult life , immunity is tightly regulated during development and early adult life, but not post-reproductively. Consistent with this, a striking decrease in expression of PMK-1 regulated genes and a decline in PMK-1 levels in aging animals was recently described , suggesting a diminished role for PMK-1 pathway in host defense towards the end of life. Therefore, a decline in immune function in late adult life may either be non-selected, or may be selected at a population level, since as discussed above, non-reproducing worms limit population numbers and stability, since they compete with their progeny for resources . The longevity of C. elegans in the wild is substantially (10-fold) shorter than under laboratory conditions ; it is probable that most worms die just after laying eggs, since nutrient availability usually is limiting in natural settings.
If the immune system of C. elegans experiences an age-related decline , which is accompanied by other age-related changes such as pharyngeal deterioration and reduced defecation , why does the bacterial load reach a strain-specific (and host-genome-specific) plateau that extends until their demise? One possibility is that a cohort effect exists, in which the fraction of worms examined in late worm adulthood constitutes a subpopulation that survived because they maintain the ability to control bacterial proliferation. Alternatively, late in life the bacterial populations develop specific syntrophic equilibria  that are resilient to changes in host milieu.
That the long-lived daf-2 mutants resist intestinal bacterial accumulation may be due to enhanced expression of luminal antimicrobial proteins and antioxidant enzymes as evidenced using DNA microarray analysis [38, 71–73]. Consistent with this hypothesis, we found that mutants lacking expression of the antimicrobial proteins lys-7 and spp-1, and the oxidative stress response enzyme ctl-2 had diminished lifespan. Since C. elegans immune responses generate ROS when bacterial pathogens are ingested , oxidative stress responses may aid in resistance by protecting against ROS-induced tissue damage. Thus, antioxidants in the gut protect from oxidative stress, preserving adequate intestinal cell function. The ctl-2 mutants also had significantly higher S. typhimurium density, consistent with an ROS resistance model. However, the intestinal bacterial densities of lys-7, lys-1, and spp-1 worms were not significantly different from N2. One explanation might be redundancy of the antimicrobial protein genes (15 encoding lysozymes and 23 encoding saposin-like domains) in C. elegans. If the numerous genes act in concert, the increased longevity of the daf-2 mutants might reflect synergies of individual genes that exert relatively small effects on lifespan and on bacterial colonization. Although the daf-2 effect also could reflect reduced senescence of the pharyngeal apparatus or defective pumping, the mixed phenotype of the daf-2;phm-2 mutant provides evidence against that hypothesis, and supports the role of enhanced expression of luminal antimicrobial proteins and antioxidant enzymes in controlling bacterial accumulation and ultimately longevity. That the colonization phenotypes of the daf-2;phm-2 double mutants is virtually identical to phm-2 early in adult life, but with aging, the daf-2 effects dominate, indicate the importance of pharyngeal function early in adult life, but that intestinal immune responses dominate as worms become senescent.
Thioredoxin expression may enhance longevity, since transgenic mice expressing human TRX-1 live longer . We confirm that trx-1 mutants have significantly decreased lifespan [47, 48], and found that intestinal bacterial density was greater in late adulthood (Additional Figure 1) when compared to N2. TRX-1 may affect C. elegans longevity and bacterial load due to its antioxidant properties , or alternately by modulation of redox-sensitive transcription factors, such as AP-1, that are activated during aging. The fact that bacterial load was greater in late adulthood is consistent with significantly enhanced expression of intestinal TRX-1 expression as worms age .
For other effectors of gut immunity, such as those encoded by dbl-1 and pmk-1, the effects on bacterial load and longevity were strongly inverse. We found that pmk-1 mutants have a shorter lifespan than previously reported . Differences in lifespan may be due to different experimental conditions. Troemel et al. added 5-fluorodeoxyuridine (FUDR) to NGM plates seeded with OP50, to prevent C. elegans progeny. However, FUDR acts to inhibit DNA synthesis, and also inhibits bacterial proliferation . That abrogating two host anti-bacterial mechanisms (e.g. dbl-1 and phm-2) produces very short survival indicates synergism between anatomical and immune defenses.
We found a strong correlation between bacterial counts and lifespan. However to better understand the biology of this host-microbial relationship, it would be critical to distinguish between continuing accumulation vs. bacterial proliferation. We address this point in a second manuscript, where we created model systems to evaluate between the possibility of bacterial persistence and proliferation or new bacterial entry . We found that host age as well as bacterial strain determine the nature of bacterial persistence in the C. elegans intestine. We also provide evidence for active competition in vivo for colonization sites as well as evidence for in vivo bacterial adaptation. We propose two mechanisms to explain the strong inverse correlation between bacterial load and lifespan. First, the intestinal milieu of older worms is more permissive for bacterial cells in general. Second, over time there is selection for bacteria that are better adapted to the intestinal niche. Our two studies provide support for both mechanisms.