High Precision Methane Measurement

Hi there, my name is Marshall McCue and today I want to introduce you to Sable Systems newest gas analyzer. It’s expanding the boundaries of metabolic phenotyping in mice and in rats. This Methane Analyzer was initially inspired by the need to explore and understand the relationship between an animal and its gut microbiome. Recent advances in gene sequencing have allowed researchers to easily characterize the community structures of the microbiome, but understanding the actual impact of the microbiome at the whole animal level offers a new horizon for research.

Just like our Stable Isotope Analyzer, this Methane Analyzer simply plugs into a Promethion system, and it can be multiplexed across as many as eight mouse or rat cages. The methane data is collected every second and automatically synchronized with all of the other behavioral and metabolic measurements that are made in real time. I’ll begin this presentation by giving a general overview of what methane is and how it’s produced in the body and discuss the implications of understanding and quantifying methane production in humans and animals. Then I’ll briefly show you inside the Methane Analyzer. And in the second half of the presentation, I’ll share some data that we’ve collected in-house using mice and rats and even humans, and present some of the emergent patterns that could be useful for phenotyping purposes.

Anyone measuring methane, or CH4, should know a few basic things about this gas. First, despite the rumors on the street, methane is actually an odorless gas. It’s a trace component in our atmosphere and it occurs at levels of about 1.9 parts per million depending on the time of the day or the season or where you are geographically. But in animals, methane is produced by anaerobic microbes; the vast majority of these inhabit the gut and make up what we call the gut microbiome. Now ruminant animals, including sheep and goats and cattle, are the best studied methane producers, and these guys can release hundreds of liters of methane per day. But Sable Systems has been building methane analyzers for these large animal applications for decades, but now we have one that can measure methane production in a single mouse.

Endogenous methane production is ubiquitous among animals beyond the mice and the rats that are going to be represented in the forthcoming graphs. Sensitive PCR testing usually focusing on ribosomal RNA sequences shows that virtually all humans also contain methanogens in their guts. But according to literature, somewhere between 20 or 60 percent, depending on who you believe, of the human population is actually considered to be methanogenic or produce a significant amount of methane. But unfortunately, from what I gather, there seems to be no clear consensus about what “significant” actually means among the research community in this context. We have a lot to learn.

From a clinical perspective, methanogenesis has been associated with gastrointestinal diseases including constipation, chronic diarrhea, malabsorption, intestinal distension, or irritable bowel syndrome (IBS) and other things. Now while antibiotic treatments can temporarily reduce the populations of methanogens, they’re not necessarily targeting the methanogenic microbes specifically, nor are they actually able to entirely eliminate these methanogenic populations, nor do we even know if that’s a good thing.

The most prevalent methanogenic archaea belong to the genera methanobrevibacter and methanosphaera, although some species of clostridium and bacteroides have the capacity to generate methane. Methanogenic microbes meet their energy demands by reducing the hydrogen gas that’s originally produced by other types of microbes that inhabit the gut, and these methanogens also compete with sulfate reducing and acetogenic microbial species for the H2 gas, and this no doubt results in real complex population dynamics that we have yet to characterize. Freshly generated methane in the body then either leaves the gut as flatus or it can be absorbed across the intestinal mucosa and enter circulation, eventually diffusing down its concentration gradient at the lung surface, not unlike exhaling carbon dioxide waste.

Let’s take a moment to look inside the Methane Analyzer, which you should never really need to open. It contains a tunable diode laser set to measure methane concentrations in a gas stream that flows constantly at a rate of somewhere between 300 and 1000 milliliters per minute. The sample chamber is maintained under a very strong vacuum, and this is bookended by a pair of high reflectivity mirrors that create an effective laser path of one kilometer within this little chamber. Ultimately, the vacuum conditions and the laser path length ensure short time constants. This means you get quick responses, and you get an unparalleled measurement resolution of down to one part per billion.

If I were you, I’d be asking these few questions. First, are the measurements stable? Well, the top graph shows a 20-minute recording at 1 Hz or one measurement per second, and it illustrates a resolution of one part per billion. You can see how the value hovers around 1.916 and 1.917 parts per million. Now if you want, this signal can be even further smoothed simply by averaging over 3-5 seconds. Second, you might be wondering “is the analyzer fast?” In the middle graph here, I’m switching among three animal cages sampled every 60 seconds. In only one of these cages I’m injecting variable amounts of methane. The steady state readings are clearly visible within the cage and reducing the dwell time from 60 seconds – even as low as 20 seconds – is acceptable. Finally, is the response linear and repeatable? Well, the bottom graph shows where I’m generating physiologically relevant methane mixtures between zero and three parts per million. What we see is it’s very linear, and it’s not shown here, but this linear response actually extends up to concentrations of 100 parts per million. Your animals might produce that much; we’ll have to see.

Although methane is typically measured in gas stream in terms of this unit parts per million, it can be useful to express the whole animal methane losses in terms of VCH4 or the rate of methane release, as I’ve done here for this mouse. This simple equation shows that VCH4 is actually the mathematical product of the fractional (this is unitless) difference between the incurrent and excurrent methane concentrations – what we call the delta FCH4 – and the flow rate of the gas as it passes through the animal chamber. As such, these units are ultimately reported in volumes per unit time, such as microliters of methane per minute for this mouse. Now while it’s technically impossible to measure methane rates of endogenous methane production inside the animal, for our purposes we can use the terms methane release and methane production interchangeably.

The mouse from the previous slide that I showed you is still shown here in blue, but here I’ve overlaid data from four other mice. Three of these mice — number 2, number 3 and number 4 – barely produced any methane and that’s not unusual in my experience. But mouse number 5 in the green produces about twice as much methane as we saw in mouse number 1. Now, this graph nicely illustrates also the two different modes of methane release; what I call continuous or constitutive release via the breath, and episodic or acute releases. So, no researchers yet have attempted to partition the source of the acute releases such as burping or flatus in mice, but it’d be easy enough to do.

In this experiment we measured a rat in a standard Promethion cage using a flow rate of four and a half liters per minute and we observed her for 72 hours (3 days). Now recall that in addition to the methane, we can simultaneously monitor lots of other physiological and behavioral parameters like movement, wheel running, energy expenditure, or food and water uptake. I’ll say we were floored when we observed this circadian pattern in methane release shown in the blue trace, because it had not yet, and it’s still as far as I know, still has not been documented yet in literature. And it turns out that the lowest methane levels seem to occur during those couple of hours in the midday periods when the rat is not eating (that’s indicated by the flat regions in the green food uptake trace). The implication of this is that the microbiome itself must be experiencing some pretty cool dynamic performance and/or community level changes in a matter of a few hours, maybe even minutes.

The previous graph was really so inspiring to us or exciting that we continued the experiment and decided to see if we could disrupt or alter the methane production of this rat in real time. So, we decided to switch from its regular chow diet to one that had substantially lower amounts of glycans and fructans in it. What we saw is within one day, the veggie treats that we call them caused the rats methane production to totally crash. Now switching the rat back to its chow diet gradually restored its methane production; interestingly though to a new even higher set point than what we saw originally. When we fasted the rat for 72 hours, this also caused the methane production to crash as we had hypothesized, but again restoring the chow diet also restored the methane production and even seems to have increased it to its highest levels yet. We really don’t have an explanation for this.

If we zoom into a specific eight-hour window during this real long 26-day recording, we can begin to see other patterns that have never been reported in literature. The most surprising of these is that the vast majority of the methane actually leaves the body in the breath, not the flatus, and we can see this in the purple trace at the bottom where the raw methane levels are reported in parts per million. You see that those lowest methane levels in the trace represent the background or ambient concentrations of methane and we can see this large constitutive methane signal during this entire period (in this eight-hour period at least). And if we zoom in even closer to a one-hour time scale, we can clearly identify the individual flatus events that only last for a few seconds as you might expect. Keep in mind that the uniquely high airflow rates of the Promethion system enable us to differentiate and even count these discrete release events. Collectively, all of these measurable phenotypic characteristics could open up new aspects of phenotyping that have not yet been considered initially or originally because of technological limitations.

Well just like the mice that I showed you earlier, humans also differ in their methane production. This is a 26-minute recording, and I asked 11 members of our research and development team to exhale a single breath near the Methane Analyzer. Clearly, person number 2 had measurably higher methane levels in their breath. Well fortunately, I’m person number 2, so I can share that with you. And just so I didn’t feel alone, I eventually found other methanogenic folks in our customer support team and our production team, and we’ve even recently hired a few more methanogenic people, but let me assure you it’s not a requirement for employment. So, this is a very simple methane screening experiment, but I wanted to dig a little deeper and find out exactly how much methane I was producing.

So, I set up a system in my bedroom where I slept in a small canopy through which 100 liters per minute of fresh air was constantly pulled, and this is the story of that night. What we see is the rate of CO2 production is shown in the blue trace and the methane production is plotted in red. What we see is during the first two hours, the rate of methane release was pretty constant; around 20 milliliters per minute. Then the traces got disturbed for a few minutes while I got up to use the restroom. However about three hours into the night we can see very clear discrete methane release events depicted as these large red spikes. Now by integrating the area under each of these spikes, we can actually quantify the amount of methane associated with every event. I hope you’d agree with me also that the majority of methane is actually leaving my body not in the flatus but rather in my breath, constitutively.

If we want to eliminate discrete methane release events and focus on the majority of that methane release that occurs in the breath, all we have to do is use a mask. So, here’s one of our research scientists wearing a mask during a one-hour measurement and air was pulled through the mask at a constant rate of 50 liters per minute. In this trial, we switched from measuring the exhaled breath to measuring the background methane or ambient methane levels, and back to the breath every six minutes. Now during the first 18 minutes is exhaled CO2, and methane relate levels remained fairly constant; who would have expected this? But then he began to exercise on a cycle ergometer for about 20 minutes and again, as you might predict, his CO2 levels increased substantially – here by fourfold. But his methane losses however seem to have decreased slightly during exercise before returning to those normal levels during the recovery period. So, do we know exactly what is happening during exercise? No. I’ve got lots of hypotheses related to exercise-induced changes in blood distribution between the gut or the skeletal muscles, or changes in blood PH, or even transcutaneous methane losses. But like I said, these are heretofore unexplored questions. I can say that one of our next steps will be to continue this exercise testing using rodent models inside our new Metabolic Treadmill. We’ll be releasing this next year – or excuse me – this year in just a couple of months and you’ll get to use that as well, so stay tuned for that.

Meanwhile, if you’d like an electronic copy of this poster that summarizes at least the rat experiment that I shared with you, please send me an email. I’m also happy to answer any specific questions about experimental applications in general so you can get off to a quick start collecting your own methane data. Meanwhile, happy phenotyping!