Stable binding of the herpes simplex virus ICP47 protein to the peptide binding site of TAP.

The herpes simplex virus (HSV) ICP47 protein inhibits the MHC class I antigen presentation pathway by inhibiting the transporter associated with antigen presentation (TAP) which translocates peptides across the endoplasmic reticulum membrane. At present, ICP47 is the only inhibitor of TAP. Here, we show that ICP47 produced in bacteria can block human, but not mouse, TAP, and that heat denaturation of ICP47 has no effect on its ability to block TAP. ICP47 inhibited peptide binding to TAP without affecting ATP binding, consistent with previous observations that the peptide binding and ATP binding sites of TAP are distinct. ICP47 bound to TAP with a higher affinity (KD approximately 5 x 10(-8) M) than did peptides, and ICP47 did not dissociate from TAP. ICP47 was not transported by TAP and remained sensitive to proteases added from the cytosolic surface of the membrane. Peptides acted as competitive inhibitors of ICP47 binding to TAP, and this inhibition required a 100- to 1000-fold molar excess of peptide. These results demonstrate that ICP47 binds to a site which includes the peptide binding domain of TAP and remains bound to this site in a stable fashion.     

Yeast “Make-Accumulate-Consume” Life Strategy Evolved as a Multi-Step Process That Predates the Whole Genome Duplication

When fruits ripen, microbial communities start a fierce competition for the freely available fruit sugars. Three yeast lineages, including baker’s yeast Saccharomyces cerevisiae, have independently developed the metabolic activity to convert simple sugars into ethanol even under fully aerobic conditions. This fermentation capacity, named Crabtree effect, reduces the cell-biomass production but provides in nature a tool to out-compete other microorganisms. Here, we analyzed over forty Saccharomycetaceae yeasts, covering over 200 million years of the evolutionary history, for their carbon metabolism. The experiments were done under strictly controlled and uniform conditions, which has not been done before. We show that the origin of Crabtree effect in Saccharomycetaceae predates the whole genome duplication and became a settled metabolic trait after the split of the S. cerevisiae and Kluyveromyces lineages, and coincided with the origin of modern fruit bearing plants. Our results suggest that ethanol fermentation evolved progressively, involving several successive molecular events that have gradually remodeled the yeast carbon metabolism. While some of the final evolutionary events, like gene duplications of glucose transporters and glycolytic enzymes, have been deduced, the earliest molecular events initiating Crabtree effect are still to be determined.     

The Effect of Attractive Interactions and Macromolecular Crowding on Crystallins Association

In living systems proteins are typically found in crowded environments where their effective interactions strongly depend on the surrounding medium. Yet, their association and dissociation needs to be robustly controlled in order to enable biological function. Uncontrolled protein aggregation often causes disease. For instance, cataract is caused by the clustering of lens proteins, i.e., crystallins, resulting in enhanced light scattering and impaired vision or blindness. To investigate the molecular origins of cataract formation and to design efficient treatments, a better understanding of crystallin association in macromolecular crowded environment is needed. Here we present a theoretical study of simple coarse grained colloidal models to characterize the general features of how the association equilibrium of proteins depends on the magnitude of intermolecular attraction. By comparing the analytic results to the available experimental data on the osmotic pressure in crystallin solutions, we identify the effective parameters regimes applicable to crystallins. Moreover, the combination of two models allows us to predict that the number of binding sites on crystallin is small, i.e. one to three per protein, which is different from previous estimates. We further observe that the crowding factor is sensitive to the size asymmetry between the reactants and crowding agents, the shape of the protein clusters, and to small variations of intermolecular attraction. Our work may provide general guidelines on how to steer the protein interactions in order to control their association.     

Continuous Solvent/Detergent Virus Inactivation Using a Packed‐Bed Reactor

Continuous virus inactivation (VI) remains one of the missing pieces while the biopharma industry moves toward continuous manufacturing. The challenges of adapting VI to the continuous operation are two‐fold: 1) achieving fluid homogeneity and 2) a narrow residence time distribution (RTD) for fluid incubation. To address these challenges, a dynamic active in‐line mixer and a packed‐bed continuous virus inactivation reactor (CVIR) are implemented, which act as a narrow RTD incubation chamber. The developed concept is applied using solvent/detergent (S/D) treatment for inactivation of two commonly used model viruses. The in‐line mixer is characterized and enables mixing of the viscous S/D chemicals to ±1.0% of the target concentration in a small dead volume. The reactor's RTD is characterized and additional control experiments confirm that the VI is due to the S/D action and not induced by system components. The CVIR setup achieves steady state rapidly before two reactor volumes and the logarithmic reduction values of the continuous inactivation process are identical to those obtained by the traditional batch operation. The packed‐bed reactor for continuous VI unites fully continuous processing with very low‐pressure drop and scalability.     

Kinetic model of ethopropazine interaction with horse serum butyrylcholinesterase and its docking into the active site.

The action of a potent tricyclic cholinesterase inhibitor ethopropazine on the hydrolysis of acetylthiocholine and butyrylthiocholine by purified horse serum butyrylcholinesterase (EC 3.1.1.8) was investigated at 25 and 37 degrees C. The enzyme activities were measured on a stopped-flow apparatus and the analysis of experimental data was done by applying a six-parameter model for substrate hydrolysis. The model, which was introduced to explain the kinetics of Drosophila melanogaster acetylcholinesterase [Stojan et al. (1998) FEBS Lett. 440, 85-88], is defined with two dissociation constants and four rate constants and can describe both cooperative phenomena, apparent activation at low substrate concentrations and substrate inhibition by excess of substrate. For the analysis of the data in the presence of ethopropazine at two temperatures, we have enlarged the reaction scheme to allow primarily its competition with the substrate at the peripheral site, but the competition at the acylation site was not excluded. The proposed reaction scheme revealed, upon analysis, competitive effects of ethopropazine at both sites; at 25 degrees C, three enzyme-inhibitor dissociation constants could be evaluated; at 37 degrees C, only two constants could be evaluated. Although the model considers both cooperative phenomena, it appears that decreased enzyme sensitivity at higher temperature, predominantly for the ligands at the peripheral binding site, makes the determination of some expected enzyme substrate and/or inhibitor complexes technically impossible. The same reason might also account for one of the paradoxes in cholinesterases: activities at 25 degrees C at low substrate concentrations are higher than at 37 degrees C. Positioning of ethopropazine in the active-site gorge by molecular dynamics simulations shows that A328, W82, D70, and Y332 amino acid residues stabilize binding of the inhibitor.     

Plant dihydroorotate dehydrogenase differs significantly in substrate specificity and inhibition from the animal enzymes

The mitochondrial membrane bound dihydroorotate dehydrogenase (DHODH; EC 1.3.99.11) catalyzes the fourth step of pyrimidine biosynthesis. By the present correction of a known cDNA sequence for Arabidopsis thaliana DHODH we revealed the importance of the very C-terminal part for its catalytic activity and the reason why--in contrast to mammalian and insect species--the recombinant plant flavoenzyme was unaccessible to date for in vitro characterization. Structure-activity relationship studies explained that potent inhibitors of animal DHODH do not significantly affect the plant enzyme. These difference could be exploited for a novel approach to herb or pest growth control by limitation of pyrimidine nucleotide pools.     

Multi-step analysis as a tool for kinetic parameter estimation and mechanism discrimination in the reaction between tight-binding fasciculin 2 and electric eel acetylcholinesterase

The mechanism of action of a potent peptidic inhibitor fasciculin 2 (Fas2) on electric eel acetylcholinesterase (eleelAChE) has been examined in a three-level analysis. Classical steps included equilibration experiments for the evaluation of high affinity binding constant and the existence of residual hydrolytic activity in a solution of completely Fas2 saturated enzyme. The two rate constants for the association (k(on)) and the dissociation (k(off)) of Fas2 with free enzyme were determined by the time course of residual enzyme activity measurements. In the third step, with a nonclassical progress curve analysis, we found that the Fas2-enzyme complex exhibited hydrolytic activity in a butyrylcholinesterase-like kinetics. The switch appears to be a consequence of steric obstruction, but also the consequence of subtle rapid conformational changes around catalytic site, upon slow single-step binding of large Fas2 molecule at the peripheral site. An unusual unilateral effect of bound Fas2 is reflected by acylation-independent association and dissociation rates and might indeed be due to inability of small acylation agent to influence the binding of a large opponent.     

Herpes Simplex Virus Type 2 ICP47 Inhibits Human TAP but Not Mouse TAP

Herpes simplex virus serotype 1 (HSV-1) expresses an immediate-early protein, ICP47, that effectively blocks the major histocompatibility complex class I antigen presentation pathway. HSV-1 ICP47 (ICP47-1) binds with high affinity to the human transporter associated with antigen presentation (TAP) and blocks the binding of antigenic peptides. HSV type 2 (HSV-2) ICP47 (ICP47-2) has only 42% amino acid sequence identity with ICP47-1. Here, we compared the levels of inhibition of human and murine TAP, expressed in insect cell microsomes, by ICP47-1 and ICP47-2. Both proteins inhibited human TAP at similar concentrations, and the K_D for ICP47-2 binding to human TAP was 4.8 × 10⁻⁸ M, virtually identical to that measured for ICP47-1 (5.2 × 10⁻⁸ M). There was some inhibition of murine TAP by both ICP47-2 and ICP47-1, but this inhibition was incomplete and only at ICP47 concentrations 50 to 100 times that required to inhibit human TAP. Lack of inhibition of murine TAP by ICP47-1 and ICP47-2 could be explained by an inability of both proteins to bind to murine TAP.Previously, we showed that herpes simplex virus serotype 1 (HSV-1) ICP47 (ICP47-1) caused major histocompatibility complex (MHC) class I proteins to be retained in the endoplasmic reticulum (ER) of cells and that antigen presentation to CD8⁺ T cells was inhibited after ICP47-1 was expressed in human fibroblasts (9). ICP47-1 blocked peptide transport across the ER membrane by TAP (2, 6), so that, without peptides, class I proteins were retained in the ER. By contrast, ICP47 did not detectably inhibit MHC class I antigen presentation in mouse cells (9) and inhibited murine TAP poorly (2, 6). ICP47-1 inhibited peptide binding to TAP without affecting the binding of ATP (1, 7) and bound with high affinity, and in a stable fashion, to human TAP (7). Peptides could competitively inhibit ICP47 binding to TAP, consistent with the hypothesis that ICP47-1 binds to a site which includes the peptide binding domain of TAP (7). Others have suggested that the present data do not exclude a distortion in TAP caused by the binding of ICP47 at a site distant from the peptide binding site (3). This seems improbable given our observations that ICP47 inhibits peptide binding and that peptides competitively inhibit ICP47 binding. In order for peptides to inhibit ICP47 binding and vice versa, one would have to invoke allosteric inhibition by both ICP47 and peptides, a highly unlikely prospect.The predicted amino acid sequence of HSV type 2 ICP47 (ICP47-2) was recently described (3), and it was of some interest that ICP47-1 and ICP47-2 share only 42% amino acid identity (see Fig. ​Fig.1A).1A). Most of the homology is near the N termini and in the central regions of the molecules. A peptide including residues 2 to 35 of ICP47-1 blocked human TAP in permeabilized cells (3). This observation was somewhat surprising given that this peptide did not include residues 33 to 51, a sequence that is most homologous between ICP47-1 and ICP47-2. Presumably, this conserved domain, and even the C-terminal third of the protein, is important in virus-infected cells for stability or for functions that are not apparent in this in vitro assay involving detergent-permeabilized cells.Open in a separate windowFIG. 1Comparison of ICP47-1 and ICP47-2 protein sequences and preparation of purified proteins. (A) The predicted amino acid sequences of ICP47-1 derived from HSV-1 strain 17 (6a) and of ICP47-2 derived from HSV-2 strain HG52 (3) are shown. The boldface, underlined letters denote identical amino acids, and the italicized letters denote conserved residues. (B) ICP47-1 and ICP47-2 were produced in Escherichia coli by expressing the proteins as GST fusion proteins by fusing the ICP47 coding sequences to GST sequences in plasmid pGEX-2T as described previously (7). Lysates from bacteria were incubated with glutathione-Sepharose and washed several times, and then ICP47-1 or ICP47-2 was eluted by incubation with thrombin, which cleaves between the GST and ICP47 sequences (7). The thrombin was inactivated with phenylmethylsulfonyl fluoride, and the ICP47 preparations were characterized by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and by Bradford protein analysis. The positions of GST-ICP47, GST, and ICP47 protein, as well as those of molecular weight markers 104, 80, 48, 34, 24, and 18 KDa in size, are indicated.Given the differences between the primary structures of ICP47-1 and ICP47-2, we were interested in whether ICP47-2 might inhibit the murine TAP. If this were the case, it would make possible animal studies of the effects of ICP47. Here, we have produced a recombinant form of ICP47-2 and compared the effects of ICP47-2 and ICP47-1 on human and murine TAP proteins expressed in insect cell microsomes. Like ICP47-1, ICP47-2 efficiently blocked human TAP but even at high concentrations did not effectively block murine TAP. Moreover, there was little or no significant binding of either protein to insect microsomes containing mouse TAP.The HSV-2 ICP47 gene was subcloned from plasmid pBB17, which contains a KpnI-HindIII 8,477-bp fragment derived from the genome of HSV-2 strain HG52 inserted into pUC19, by using PCR to amplify ICP47-2 coding sequences. One PCR primer hybridized with the 5′ end of the ICP47-2 coding sequences and extended 5′ to generate a new BglII site just upstream of the initiation codon. The second PCR primer hybridized with 3′ sequences of the ICP47-2 gene, then diverged to produce an EcoRI site just downstream of the translation termination codon. After PCR, the DNA fragment was digested with EcoRI and inserted into the HincII (blunt) and EcoRI sites of pUC19, producing plasmid pUC47-2, which was subjected to DNA sequencing. The ICP47-2 coding sequences were excised from pUC47-2 with BglII and EcoRI and inserted into the BamHI and EcoRI sites of pGEX-2T to generate a fusion protein with glutathione S-transferase (GST). The ICP47-GST fusion protein was expressed in bacteria and purified by using glutathione-Sepharose, and then the GST sequences were removed with thrombin as described previously for ICP47-1 (7). A comparison between the predicted amino acid sequences of ICP47-2 and ICP47-1 is shown in Fig. ​Fig.1,1, with a comparative gel (Fig. ​(Fig.1B)1B) showing the purified preparations of ICP47-1 and ICP47-2 from bacteria. Microsomes purified from Sf9 insect cells infected with baculoviruses expressing human TAP1 and TAP2 have been described previously (7, 8), as were microsomes from Drosophila cells expressing murine TAP1 and TAP2 (1). We previously estimated that approximately 2% of the protein associated with the insect microsomes was human TAP (7), and the microsomes containing mouse TAP possessed similar TAP activity (see below). Peptide translocation by these microsomes was measured by using a library of ¹²⁵I-labelled peptides (5) that are glycosylated after transport into the ER. Radioactive peptides able to bind to concanavalin A were quantified as an indirect measure of peptide transport (6). Over a range of membranes from 2.5 to 20 μl, with protein concentrations of 10 to 12 mg/ml for human TAP microsomes and 5.0 to 7.0 mg/ml for mouse TAP microsomes, there was a linear increase in peptide transport (Fig. ​(Fig.2).2). Thus, peptides and ATP were not limiting. Peptide transport was specific because the transport observed with control membranes not containing TAP amounted to less than 1% of that observed when microsomes contained TAP. The levels of peptide transport associated with microsomes containing human or mouse TAP were also compared and standardized. Thus, in subsequent assays, 7.5 to 10 μl of microsomes exhibiting similar amounts of TAP activity were used. Open in a separate windowFIG. 2Peptide transport by insect microsomes containing human or murine TAP. Microsomes were derived from insect Sf9 cells coinfected with BacTAP1 and BacTAP2 (Human TAP) (7) or from Sf9 cells infected with a control baculovirus, BacgH (Human control). Alternatively, microsomes were derived from Drosophila cells induced to express mouse TAP (Murine TAP) (1) or from Drosophila cells which were not induced to express mouse TAP (Murine control). Various concentrations of each microsome preparation were incubated with ¹²⁵I-labelled peptides and 5 mM ATP in a volume of 150 μl for 10 min at 23°C. The microsomes were washed, pelleted, and disrupted in detergent as described previously (7). Peptides able to bind to concanavalin A-Sepharose were eluted with alpha-methylmannoside and quantified (7).ICP47-2 inhibited peptide transport by human TAP, and the inhibition was similar to that of ICP47-1; the 50% inhibitory concentration (IC₅₀) for ICP47-2 was 0.24 μM and for ICP47-1 was 0.27 μM (Fig. ​(Fig.3A).3A). In other experiments the IC₅₀ values for ICP47-1 and ICP47-2 varied from 0.15 to 0.35 μM, and there were no experiments in which there was a significant difference in the abilities of the two proteins to inhibit human TAP. Moreover, the binding properties of ICP47-2 to human TAP were similar to those of ICP47-1. Binding experiments were performed as described previously for ICP47-1 (7) by using membranes containing human TAP and ¹²⁵I-labelled ICP47-2. Specific binding of ICP47-2 was calculated by subtracting the binding to control microsomes derived from insect cells infected with a baculovirus expressing HSV gH (7). The binding of ICP47-2 was saturable, so that at a protein concentration of 1 μM approximately 16 ng of protein bound to human TAP (Fig. ​(Fig.4A).4A). In previous experiments with a similar preparation of insect microsomes containing human TAP, the binding of ICP47-1 also saturated at 15 to 16 ng (7). The ICP47-2 binding data were analyzed in a standard Scatchard plot, and the K_D was calculated to be 4.8 × 10⁻⁸ M (Fig. ​(Fig.4B),4B), compared with 5.2 × 10⁻⁸ M for ICP47-1 (7). These values are greater than those of high-affinity peptides that bind to human TAP with affinities reaching 4 × 10⁻⁷ M, though the vast majority of peptides bind to TAP with much lower affinities (8). Open in a separate windowFIG. 3Inhibition of human and murine TAP-mediated peptide transport by ICP47-1 and ICP47-2. TAP assays were performed as described in the legend for Fig. ​Fig.22 by using insect microsomes containing human TAP (10 μl of membranes containing 12 mg of membrane protein per ml) (A) or murine TAP (7.5 μl of membranes containing 4.8 mg of membrane protein per ml but with equivalent levels of TAP activity compared with microsomes containing human TAP) (B) and various concentrations of ICP47-1 and ICP47-2. The results shown are combined from two separate experiments, each involving human and murine TAP.Open in a separate windowFIG. 4Binding of ICP47-2 to human TAP. (A) Microsomes (15 μl of membranes with a 7.5-mg/ml concentration of membrane protein) derived from Sf9 cells expressing TAP1 and TAP2 or expressing HSV-1 gH (control membranes not containing TAP) were incubated with various amounts of ¹²⁵I-labelled ICP47-2 for 60 min at 4°C as described previously (7). Binding to control membranes was subtracted from binding to microsomes containing TAP at each point. (B) Scatchard analysis of the data in panel A. The K_D for ICP47-2 binding to TAP was calculated to be 4.8 × 10⁻⁸ M.To determine whether ICP47-2 could inhibit the murine TAP, microsomes from insect cells expressing mouse TAP were incubated with various concentrations of ICP47-1 and ICP47-2 and TAP assays were performed. Inhibition of the mouse TAP was observed with both ICP47-1 and ICP47-2, but relatively high concentrations of both proteins were required (Fig. ​(Fig.3B).3B). The IC₅₀ values for ICP47-1 and ICP47-2 in this experiment were 10.8 and 16.2 μM, respectively. However, we were unable to reduce TAP activity beyond approximately 40% with ICP47-1 or ICP47-2 concentrations reaching 30 μM. This was 100 times the concentration required to inhibit human TAP by 50%. We attempted to measure the specific binding of radiolabelled ICP47-1 and ICP47-2 to microsomes containing mouse TAP in experiments similar to those shown in Fig. ​Fig.4.4. However, there was little specific binding of ICP47-1 and ICP47-2, and it was difficult to measure binding at lower protein concentrations. We therefore measured binding at a single, higher protein concentration (2.75 μM), one sufficient to inhibit 10 to 20% of the mouse TAP activity and all of the human TAP activity. In this experiment, specific binding to microsomes containing murine TAP was determined by subtracting the binding to microsomes from insect cells that were not induced to express murine TAP (1). The binding of ICP47-1 and ICP47-2 to human TAP was easily measured (Fig. ​(Fig.5),5), although under these conditions it is important to note that ICP47-1 and ICP47-2 were present at concentrations beyond those required to saturate the TAP (Fig. ​(Fig.4A).4A). By contrast, it was found that there was little or no significant binding of ICP47-1 or ICP47-2 to microsomes containing murine TAP when background binding to control membranes was subtracted. In the experiment shown, specific ICP47-2 binding was greater than zero, but in other experiments this binding was less than zero, and thus we concluded that there was no detectable binding overall. In every experiment, it was clear that the level of binding of ICP47-1 and ICP47-2 to murine TAP was at least 25-fold lower than to human TAP. However, the human TAP present in these microsomes was limiting in these experiments, and thus it is very likely that the 25-fold difference between the levels of binding to human and mouse TAP is an underestimate. More likely this difference is 50- to 100-fold. On the basis of the inhibitory concentrations required to block murine TAP and the binding studies described above, estimates of the binding affinities of ICP47-1 and ICP47-2 for murine TAP may fall in the range of 5 × 10⁻⁶ M. Therefore, ICP47-1 and ICP47-2 bind poorly to the murine TAP, and this largely accounts for their inability to block mouse TAP peptide transport. Open in a separate windowFIG. 5Binding of ICP47-1 and ICP47-2 to microsomes containing murine TAP. Microsomes containing human TAP or control membranes without human TAP (100 μg of membrane protein per 150-μl assay) or microsomes containing mouse TAP or control membranes without mouse TAP (50 μg of membrane protein with the same TAP activity as with the human microsomes) were incubated with ¹²⁵I-labelled ICP47-1 or ICP47-2 at 2.75 μM for 60 min at 4°C. The microsomes were washed twice, pelleted, and disrupted with detergents as described previously (7). Radioactivity associated with the microsomes was quantified by gamma counting. “ICP47 bound” refers to specific binding, calculated by subtracting the binding to control membranes (without TAP) from that observed with microsomes containing human or murine TAP.In summary, ICP47-2 and ICP47-1 could block human TAP and bound to TAP with similar high affinities. It was interesting that these two proteins, whose primary structures are only about 40% identical, inhibit human TAP with indistinguishable profiles and bind to human TAP with virtually identical affinities. Moreover, both proteins blocked murine TAP poorly and only at high protein concentrations and could not bind to murine TAP. These results, at face value, would suggest that mice will not be an appropriate model in which to test the effects of ICP47 on HSV replication or as a selective inhibitor of CD8⁺ T-cell responses in other systems. However, we recently found that an HSV-1 ICP47 mutant showed dramatically reduced neurovirulence in mice, without altering the course of disease in the cornea (4). Therefore, ICP47 may attain sufficient concentrations in certain cells in the nervous systems of mice to inhibit TAP. This may be related to the fact that TAP and class I proteins are expressed at low levels in the nervous system. Alternatively, ICP47 may have other functions in the nervous system.     

A Study on the Fundamental Mechanism and the Evolutionary Driving Forces behind Aerobic Fermentation in Yeast

Baker’s yeast Saccharomyces cerevisiae rapidly converts sugars to ethanol and carbon dioxide at both anaerobic and aerobic conditions. The later phenomenon is called Crabtree effect and has been described in two forms, long-term and short-term effect. We have previously studied under fully controlled aerobic conditions forty yeast species for their central carbon metabolism and the presence of long-term Crabtree effect. We have also studied ten steady-state yeast cultures, pulsed them with glucose, and followed the central carbon metabolism and the appearance of ethanol at dynamic conditions. In this paper we analyzed those wet laboratory data to elucidate possible mechanisms that determine the fate of glucose in different yeast species that cover approximately 250 million years of evolutionary history. We determine overflow metabolism to be the fundamental mechanism behind both long- and short-term Crabtree effect, which originated approximately 125–150 million years ago in the Saccharomyces lineage. The “invention” of overflow metabolism was the first step in the evolution of aerobic fermentation in yeast. It provides a general strategy to increase energy production rates, which we show is positively correlated to growth. The “invention” of overflow has also simultaneously enabled rapid glucose consumption in yeast, which is a trait that could have been selected for, to “starve” competitors in nature. We also show that glucose repression of respiration is confined mainly among S. cerevisiae and closely related species that diverged after the whole genome duplication event, less than 100 million years ago. Thus, glucose repression of respiration was apparently “invented” as a second step to further increase overflow and ethanol production, to inhibit growth of other microbes. The driving force behind the initial evolutionary steps was most likely competition with other microbes to faster consume and convert sugar into biomass, in niches that were semi-anaerobic.