Macromolecular crystals for X-ray diffraction studies are typically grownin vitrofrom pure and homogeneous samples; however, there are examples of protein crystals that have been identifiedin vivo. Recent developments in micro-crystallography techniques and the advent of X-ray free-electron lasers have allowed the determination of several protein structures from crystals grownin cellulo. Here, an atomic resolution (1.2 Å) crystal structure is reported of heterogeneous milk proteins grown inside a living organism in their functional niche. Thesein vivo-grown crystals were isolated from the midgut of an embryo within the only known viviparous cockroach,Diploptera punctata. The milk proteins crystallized in space groupP1, and a structure was determined by anomalous dispersion from the native S atoms. The data revealed glycosylated proteins that adopt a lipocalin fold, bind lipids and organize to form a tightly packed crystalline lattice. A single crystal is estimated to contain more than three times the energy of an equivalent mass of dairy milk. This unique storage form of nourishment for developing embryos allows access to a constant supply of complete nutrients. Notably, the crystalline cockroach-milk proteins are highly heterogeneous with respect to amino-acid sequence, glycosylation and bound fatty-acid composition. These data present a unique example of protein heterogeneity within a singlein vivo-grown crystal of a natural protein in its native environment at atomic resolution.

Copper-dependent lytic polysaccharide monooxygenases (LPMOs) classified in Auxiliary Activity (AA) families are considered indispensable as synergistic partners for cellulolytic enzymes to saccharify recalcitrant lignocellulosic plant biomass. In this study, we characterized two fungal oxidoreductases from the new AA16 family. We found that MtAA16A from Myceliophthora thermophila and AnAA16A from Aspergillus nidulans did not catalyze the oxidative cleavage of oligo- and polysaccharides. Indeed, the MtAA16A crystal structure showed a fairly LPMO-typical histidine brace active site, but the cellulose-acting LPMO-typical flat aromatic surface parallel to the histidine brace region was lacking. Further, we showed that both AA16 proteins are able to oxidize low-molecular-weight reductants to produce H₂O₂. The oxidase activity of the AA16s substantially boosted cellulose degradation by four AA9 LPMOs from M. thermophila (MtLPMO9s) but not by three AA9 LPMOs from Neurospora crassa (NcLPMO9s). The interplay with MtLPMO9s is explained by the H₂O₂-producing capability of the AA16s, which, in the presence of cellulose, allows the MtLPMO9s to optimally drive their peroxygenase activity. Replacement of MtAA16A by glucose oxidase (AnGOX) with the same H₂O₂-producing activity could only achieve less than 50% of the boosting effect achieved by MtAA16A, and earlier MtLPMO9B inactivation (6 h) was observed. To explain these results, we hypothesized that the delivery of AA16-produced H₂O₂ to the MtLPMO9s is facilitated by protein-protein interaction. Our findings provide new insights into the functions of copper-dependent enzymes and contribute to a further understanding of the interplay of oxidative enzymes within fungal systems to degrade lignocellulose.

The recently discovered lytic polysaccharide monooxygenases (LPMOs) are Cu-containing enzymes capable of degrading polysaccharide substrates oxidatively. The generally accepted first step in the LPMO reaction is the reduction of the active-site metal ion from Cu²⁺ to Cu⁺. Here we have used a systematic diffraction data collection method to monitor structural changes in two AA9 LPMOs, one from Lentinus similis (LsAA9_A) and one from Thermoascus aurantiacus (TaAA9_A), as the active-site Cu is photoreduced in the X-ray beam. For LsAA9_A, the protein produced in two different recombinant systems was crystallized to probe the effect of post-translational modifications and different crystallization conditions on the active site and metal photoreduction. We can recommend that crystallographic studies of AA9 LPMOs wishing to address the Cu²⁺ form use a total X-ray dose below 3 × 10⁴ Gy, while the Cu⁺ form can be attained using 1 × 10⁶ Gy. In all cases, we observe the transition from a hexacoordinated Cu site with two solvent-facing ligands to a T-shaped geometry with no exogenous ligands, and a clear increase of the θ₂ parameter and a decrease of the θ₃ parameter by averages of 9.2° and 8.4°, respectively, but also a slight increase in θ_T. Thus, the θ₂ and θ₃ parameters are helpful diagnostics for the oxidation state of the metal in a His-brace protein. On binding of cello-oligosaccharides to LsAA9_A, regardless of the production source, the θ_T parameter increases, making the Cu site less planar, while the active-site Tyr—Cu distance decreases reproducibly for the Cu²⁺ form. Thus, the θ_T increase found on copper reduction may bring LsAA9_A closer to an oligosaccharide-bound state and contribute to the observed higher affinity of reduced LsAA9_A for cellulosic substrates.

Lytic Polysaccharide Monooxygenases (LPMOs) oxidatively cleave recalcitrant polysaccharides. The mechanism involves (i) reduction of the Cu, (ii) polysaccharide binding, (iii) binding of different oxygen species, and (iv) glycosidic bond cleavage. However, the complete mechanism is poorly understood and may vary across different families and even within the same family. Here, we have investigated the protonation state of a secondary co-ordination sphere histidine, conserved across AA9 family LPMOs that has previously been proposed to be a potential proton donor. Partial unrestrained refinement of newly obtained higher resolution data for two AA9 LPMOs and re-refinement of four additional data sets deposited in the PDB were carried out, where the His was refined without restraints, followed by measurements of the His ring geometrical parameters. This allowed reliable assignment of the protonation state, as also validated by following the same procedure for the His brace, for which the protonation state is predictable. The study shows that this histidine is generally singly protonated at the Nε2 atom, which is close to the oxygen species binding site. Our results indicate robustness of the method. In view of this and other emerging evidence, a role as proton donor during catalysis is unlikely for this His.