Unfortunately, low-cost sequencing was unavailable at that time, so the study required post-preservation plating of organisms to distinguish, which severely limited the number of organisms that could be studied. With such little information for comparison, it is difficult for us to infer whether the patterns observed in the preservation experiments originated simply from the differential responses of oral microbes or from some mutually beneficial relationships among these microbes. Our plan to investigate composition-based differential responses to preservation will certainly help discriminate between these two sources, but we may need to supplement our planned approach with experiments on the viability of single organisms.Figure 32 outlines the experimental scheme for this phase. All organisms were purchased commercially from either ATCC or DMSZ as sealed plates or pellets in sealed ampules. Compositions of the inocula were designed based on the results from the preservation experiments, where we observed a dichotomy between the Veillonella and Streptococcus taxa and a simultaneous possible affiliation between the Veillonella and Prevotella taxa. Controls 1 and 2 represent the two compositional boundaries, one with a high abundance of the Veillonella dispar and the other with a high abundance of Streptococcus salivarius. The other four compositions were constructed as a gradient between the two boundaries, blueberry container size with the proportion of Streptococcus oralis held constant so that the effect of Prevotella salivae on the community can be studied.
These six compositionally defined microcosms will be anaerobically incubated for 72 hours at 37°C in conditions identical to those in the preservation experiment, refrigerated for one day with minimal physical, chemical and biological disturbance, and harvested without propagation. Harvested cells will be extracted with the PowerSoil kit and quantified with the PicoGreen method detailed in the Section 4.2.2. After quantification, we will perform qPCR on the samples with organism-specific rpoB primers to quantify the number of cells from each strain. We will also perform qPCR on cultures before preservation to gain an understanding of how the “baseline” composition changes in response to refrigeration.All organisms were cultured in sealed tubes with sterile SHI medium at pH 7.0, which consisted of proteose peptone, trypticase peptone, yeast extract, KCl, and pig mucin III, as well as low concentrations of menadione, L-arginine, urea, hemin, and Nacetylmuramic acid. Components that were robust against high heat and high pressure were added to the media prior to autoclaving; components sensitive to heat or pressure, such as sucrose and sheep blood, were added after autoclaving, by passing through 0.2µm filters. Sterile and complete media was degassed by passing filtersterilized anaerobic gas through the media for 15 or more minutes. Single colonies of organisms were selected with colony pickers and gently resuspended in degassed medium in a culture tube. The tube was then sealed with a rubber cap, flushed and filled with the aforementioned anaerobic gas, and incubated at 37°C without shaking. Approximately every 24 hours, we briefly flushed and refilled the tubes with anaerobic gas to maintain positive pressure.
The growth of the organisms was periodically monitored by optical density measurements at 600nm, and growth curves constructed were based on these measurements. After organisms reached their growth phases, we inoculated blood-agar plates with loopfuls of the liquid cultures and incubated the plates at 37°C in a sealed chamber under an anaerobic atmosphere. All genomic DNA used in this phase of the project was extracted with the PowerSoil DNA Isolation kit according to manufacturer’s instructions.For each organism, two sets of rpoB primers with Tm values between 58°C and 62°C and PCR product length between 100bp and 170bp were designed in silico. Table 3 details the characteristics and sequences of the forward and reverse primers. All sequences are written 5’ to 3’. “Name” represents the name of the primer, an abbreviation of the strain name plus a number; temperatures are in units of degrees Celsius; “PL” is the PCR product length, in units of base pairs. The primer names will be used in subsequent visualizations of results. Upon receiving the primers, we dissolved them in water and determined the concentrations using the Qubit 2.0 instrument with the high-sensitivity single-stranded DNA assay kit . We then performed colony PCR in duplicate, one colony per reaction, to test the efficiencies of the primers. For each PCR reaction, we use a total volume of 20µL: 1µL each of the forward and reverse primers, 1µL extracted gDNA or colony, and 17µL SsoAdvanced Universal SYBR Green Supermix . We used the following general PCR protocol: initial step at 95°C for 2 minutes, then 35 cycles consisting of 30s at 95°C, 30s, 30s of gradient temperature, and 60s at 72°C, followed by extension at 72°C for 10 minutes.
For the P. salivae PCR, we used 45s instead of 60s for the 72°C step, and for the V. dispar PCR, we used the same protocol as for P. salivae but with a lower cycle number of 30. After selecting the better primer set for each organism, we optimized the annealing temperature for that organism and checked for crosstalk between the primer set and the three off-target organisms. We also attempted to reduce the P. salivae cycle number. PCR products were mixed with gel loading dye, denatured at 95°C for 3 minutes, and loaded onto 8% urea gels. Gels were run at 215V, 100V, or a mix of the two until the fastest markers in the ladders almost reached the bottom. For more accurate reference, we used both the Low Molecular Weight DNA Ladder and the TriDye Ultra Low Range ladder . Gels were stained for 20 – 30 minutes in SYBR Gold diluted 1:10,000 in 1X TBE buffer, and then visualized using short-range ultraviolet Radiation.The growth curves of three of the four organisms in liquid SHI under anaerobic atmosphere are shown in Figure 33, as estimated by optical density measurements at 600nm. Each data point represents the average of OD600 values for two tubes. The baseline of the optical measurement consisted of a culture tube with degassed sterile SHI medium, incubated at 37°C under anaerobic conditions simultaneously with the inoculated media. Unlike conventional bacterial growth curves, these curves appeared somewhat jagged, with a dip around 13.5 hours and another dip around 34.5 hours. These dips did not fully correspond to the intermittent flushing and refilling of the tubes with anaerobic gas, which occurred approximately every 24 hours. In addition, all three organisms grew somewhat slowly, and their growths seemed to plateau starting around the 25-hour mark. It is possible that the nutrients in the media had been depleted at 25 hours, though we would need to extend the incubation time for a more complete curve to be sure. Interestingly, none of the curves showed a lag phase at the start of incubation. With the time available, we were unable to grow Veillonella dispar successfully.Accurate predictions of time-to-death are important in bio-medicine to allow patients to better consider their future, for health professionals to make more informed medical decisions, and for family members to have realistic expectations . Consequently a robust literature exists in the medical sciences concerned with predicting survival time of persons suffering from one or more of any number of different fatal diseases . This literature extends to the veterinary sciences and basic biology, as well to papers on predicting survival times for terminally-ill pets including dogs , cats and laboratory rodents . One of the common threads across this than atological research on humans, pets and rodents is that the predictions are informed, growing raspberries in container not only by extensive knowledge of the causes, origins and natural courses of the underlying diseases but, at least for humans, also by access to deep databases containing the outcome of tens of thousands, if not millions, of previous cases at different disease stages that end in the individual’s death. Remarkably, even with access to this extensive information on disease progress and outcomes, the vast majority of models in bio-medicine used for predicting survival times are both imprecise and inaccurate, with short-term survival usually overestimated and long-term survival often underestimated .
The inaccuracy of sophisticated models designed to predict human survival time that are based on extraordinary amounts of data and extensive biomedical information on both the disease and the patient lays bare the challenges of predicting survival time in non-human species for which virtually none of this information available. Inasmuch as there is no practical need for developing models for predicting time to death in the overwhelming majority of non-domesticated organisms, it is unsurprising that the literature in this area is extraordinarily scarce. With the exception of the papers by Rauser, Mueller and their colleagues in which these researchers classify Drosophila melanogaster females according to whether or not they were in a death spiral stage , and the paper by Papadopoulos and his colleagues who showed that supine behavior was predictive of impending death in medflies, we are aware of no other papers that are expressly concerned with predicting the timing of death in fruit flies or other non-domestic group of organisms.However, there are two groups of studies whose results are related to our research. The first group involves end-of-life egg laying patterns in the context of aging including: Novoseltzev and his colleagues on the senescent stage of D. melanogaster and the medfly as exponentially-decreasing rate of reproduction, Curtsinger and his colleagues on working versus retired who characterized the end-phase degree of “roughness” of individual egg laying using the fractal concept of lacunarity ; and Rogina and her colleagues who, by manipulating the timing of mating in D. melanogaster females, discovered reproductive patterns that were conditional on when females mated. All experiments in her and her colleague’s studies revealed characteristics suggesting that longer-lived flies passed through three stages, the last of which they labeled “declining terminal”. The second group of studies related to our work, albeit more tangentially, includes research concerned with the timing of reproduction relative to their death. One study representative of this group involves research concerned with the length of remaining life in individual flies relative to others based on the rate of decrease in reproduction after the peak. For example, Müller and his colleagues showed that the exponential rate of decrease in egg laying by female medflies predicted the remaining life spans of individuals. The other area within this second group involves research on the cost of reproduction in which increments of reproduction in early life result in decrements in survival later in life including papers by Harshman and Zera on mechanisms and modeling papers by Mangel and Heimpel , Fletcher and his coworkers and Rosenheim concerned with foraging strategies relative to both remaining reproduction and lifespans. There are at least three reasons why developing predictive models of impending death or that can be used to distinguish terminal periods in non-human species such as fruit flies is potentially important. First, accurate predictive models can affirm patterns associated with the transition from early stages of aging to the late stages in which the manifestations of increasing frailty are catastrophic, which is to say, ends in death. This same “staging” process might have the potential to be used to identify parse other stages of the aging process. Second, predictive models for death could be used in intervention trials in which flies showing early signs of impending death could be identified and subjected to treatments designed to test death-postponing interventions. Third, the model-building required to identify different patterns of reproduction that identify impending death in flies, in some form, may be relevant to the model-building process predicting the timing of death in humans. Indeed the model-building process itself might shed light on the classification process involved in separating the non-dying and dying as well as in calibrating survival time . In light of the paucity of information on end-of-life patterns of life history phenomena in non-human species as well as the potential importance to basic biology of these types of studies, the overarching goal of our study was to determine whether it was possible to distinguish across three fruit fly species a set of general patterns of egg laying in individual females who were near death from the egg-laying patterns in individual who were not near death . Our specific aims were to: Visualize and summarize the egg-laying patterns in individuals from all species with respect to both their chronological and thanatological ages; create subsets of these data in containing egg laying sequences from both the end-of-life and mid-life ; and test hypotheses that the patterns of egg laying between the two categories of egg-laying segments have distinct patterns.