Diligent readers of this blog will remember my earlier post on making rice wine. While enjoying the fruits of our labor (hic!), I was struck by how sweet and fruity it was, despite being made from glutinous rice, which is essentially just starch. Glutinous rice differs from normal rice in having more amylopectin, which is branched, than the linear amylose. My colleague Erik P runs samples regularly on the gas chromatograph-mass spectrometer (GC-MS), so I convinced him to sneak in a sample of the wine too.
I thought we would be able to find some esters or aldehydes, typical flavor molecules (esters are used as artificial fruit aroma, for example), but these volatile small molecules were not detected with the preparation method he used. Instead we found mostly the organic acids and sugars. And this gave us a surprise.
The highest peak in the GC belonged to glucose. No prizes for guessing, because starch is basically just repeated glucose units. As the yeast digests the starch it chops it up to glucose monomers.
Second highest, though, was lactose. There was definitely no milk added (it was not rice pudding), so where did the lactose come from? Lactose is a disaccharide, i.e. it is made up of two simpler sugar units, namely glucose and galactose. Galactose is similar to glucose, having the same molecular formula, but is a stereoisomer. Simply cleaving off disaccharide units from starch, though, would yield maltose, which is made of two glucose units. However, we did not detect maltose. My best guess at the moment is that the yeast is converting one of the glucose units to galactose, using some kind of epimerase, which are known from yeast (reference: some cursory googling).
Erik had a nice hypothesis for why this might be a useful strategy for the yeast. If its competitors in the natural environment are unable to use lactose, then converting the immediate digestion products of starch to lactose would let the yeast hoard the sugar reserves in plain sight. The breakdown of starch has to go extracellularly, and glucose is readily taken up by virtually all heterotrophic organisms. Mixing things up by scrambling some of them to galactose would allow only those who had the right enzymes to consume the mobilized resources.
Coincidentally we had heard just some days ago a nice talk from a visiting scientist on a similar strategy used by a certain bacterial strain to break down extracellular polysaccharides of (ironically) yeast. The bacterium breaks down the polysaccharide yeast mannan outside the cell, but only partially. Other bacteria mostly lack the enzymes to digest the partial breakdown product, so it can transport the partial product into itself at leisure for complete digestion.
I am a bit worried about finding lactose in my rice wine though, because lactose intolerance is a problem that affects many people. If we had not analyzed this sample, I would have not expected to find appreciable quantities of lactose in a vegan rice product, of all things. Who knows what other microbially-mediated beverages and food might upset the lactose intolerant?
As a public service, we should run samples of various alcoholic beverages on the GC-MS to check, but the machine only needs milligram quantities: What should we do with the leftovers…? (Hic!)