*4.1. Macromolecule Production*

The coding sequence of CfAM for the A, B and C domains was amplified from *Cordyceps farinosa* gDNA by the polymerase chain reaction (PCR). The PCR fragment was obtained using primer pairs: 5 -ACACAACTGGGGATCCACCATGAAGCTTACTGCGTCCCTC-3 and 5 -GATGGTGATGGGATCCTTACTGCGCAACAAAAACAATGGG-3 . The fragment was then ligated in the expression vector pSUN515 using *Bam*HI and *Xho*I restriction sites. The ligation protocol was performed according to the IN-FUSION™ Cloning Kit instructions. A transformation of TOP10 competent *E. coli* cells (Tiangen, Beijing China) with the plasmid, containing the CfAM gene, was performed and positive clones confirmed by sequencing. The transformation of *Aspergillus oryzae* (strain *MT3568*) with the expression vector comprising CfAM gene was performed according to patent application WO95/002043 [33]. After incubation for 4–7 days at 37 ◦C, spores of four transformants were inoculated into 3 mL of YPM medium. After 3-day cultivation at 30 ◦C, the culture broths were analyzed by SDS-PAGE to identify the transformant producing the largest amount of recombinant mature amylase with an estimated size of 48 kDa. Spores from the best expressing transformant were cultivated in YPM medium in shake flasks for 4 days at a temperature of 30 ◦C. The culture broth was harvested by filtration using a 0.2 μm filter device, and the filtered fermentation broth was used for purification and further assays.

RpAM was cloned and expressed in a similar manner as CfAM while TeAM was expressed in *Pichia pastoris* with a similar protocol to that described for the lipase from *Gibberella zeae* [34]. The entire coding sequence of TeAM was amplified from cDNA by the polymerase chain reaction and transformation into ElectroMax DH10B competent cells (Invitrogen, Waltham, MA, USA) by electroporation. Transformed cells were plated on LB plates containing 100 mM ampicillin. After overnight incubation at 27 ◦C, a positive clone was selected by colony PCR and confirmed by sequencing. The plasmid DNA of the positive clone was linearized with PmeI (NEB, Ipswich, MA, USA) and transformed into *Pichia pastoris* KM71 (Invitrogen, Waltham, MA, USA) following the manufacturer's instructions. An amylase positive clone was inoculated into 3 mL buffered minimal sorbitol complex medium and incubated at 28 ◦C for 3 days until the OD600 reached 20. Methanol was added to the culture daily to a final concentration of 0.5% for the following 4 days. On day 4 of induction, the culture supernatant was separated from the cells by centrifugation and the pH of the supernatant was adjusted to 7.0.

The CfAM culture broth was precipitated with ammonium sulphate (80% saturation), then dialyzed with 20 mM Na-acetate at pH 5.0. The solution was loaded on to a Q Sepharose Fast Flow column (GE Healthcare, Brondby, Denmark) equilibrated with 20 mM Na Acetate at pH 5.0. Protein was eluted with a salt gradient from zero to 1 M NaCl Fractions were analyzed for amylase activity and pooled accordingly. The flow-through fraction, containing the bulk of amylase activity was supplemented with ammonium sulphate to a final concentration of 1.2 M and then loaded on to Phenyl Sepharose 6 Fast Flow column (GE Healthcare, Brondby, Denmark). The activity was eluted by a linear gradient of decreasing salt concentration. The fractions with activity were analyzed by SDS-PAGE and then concentrated for further use.

Amylase activity was detected by Azo dyed and azurine cross-linked hydroxyethyl-amylose (AZCL-HE-amylose) (Megazyme International Ireland Ltd., Bray, Ireland) as substrate. In addition, 10μL enzyme sample and 120 μL 0.1% substrate at pH 7 were mixed in a microtiter plate and incubated at 50 ◦C for 30 min. Then, 70 μL supernatant was transferred to a new microtiter plate and the absorption at 595 nm determined. All reactions were done as duplicates.

#### *4.2. Biochemical Characterisation*

#### 4.2.1. pH Optimum

To determine the pH Optimum, each enzyme (3 μL of a 0.5 mg/mL solution) was incubated with 40 μL 1% substrate (AZCL-HE-amylose) (Megazyme International Ireland Ltd., Bray, Ireland). The pH between 2 and 11 was adjusted using 100 μL of B&R buffer (Britton–Robinson buffer: 0.1 M boric acid, 0.1 M acetic acid, and 0.1 M phosphoric acid, adjusted to pH-values 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0, 10.0 and 11.0 with HCl or NaOH) [35]. The reactions were incubated at 30 ◦C for 30 min and afterwards 60 μL were transferred in a new microtiter plate and the absorption was measured at 595 nm.

#### 4.2.2. Temperature Optimum

To determine the Temperature Optimum, each enzyme was incubated with 100 μL 0.1% substrate (AZCL-HE-amylose) (Megazyme International Ireland Ltd., Bray, Ireland) in 50 mM Na Acetate pH 4.3. The substrate solution was preincubated at 20–90 ◦C for 5 min and the reaction was started by addition of 3 μL of enzyme solution (0.5 mg/mL). The reaction mixture was further incubated at the respective temperature for 30 min at 950 rpm. The reaction was stopped by rapid cooling on ice. Afterwards, 60 μL of each reaction was transferred in a microtiterplate and the absorption was measured at 595 nm. Each reaction was performed in triplicate.

#### 4.2.3. Product Profile

For product profile determination, each enzyme (15 μL) was incubated with 120 μL 0.1% substrate (AZCL-HE-amylose) (Megazyme International Ireland Ltd., Bray, Ireland) at pH 5 and 62 ◦C for 14 h. 70 μL of each reaction was mixed with equal amounts of Acetonitril. The mixture was centrifuged for 30 min at 16.000× *g* and the supernatant was analyzed using HPAEC with pulsed amperometric detection.

#### *4.3. Crystallisation*

#### 4.3.1. RpAM

The concentrated protein was mixed with acarbose in a molar ration of 4:1 before the initial screening in 96 well format using commercially available screens. An initial hit (0.2 M NaCl, 0.1 M Na-acetate pH 4.6, 30% MPD) was further refined in 24-well format using the initial crystals as seeds. Crystals suitable for data collection were cryoprotected using 25% glycerol and flash frozen in liquid nitrogen prior data collection.

#### 4.3.2. TeAM

Prior to providing the sample to York, the protein was deglycosylated using Endo-H treatment. The protein was concentrated using Amicon (Merck, Germany) filter units and stored at −80 ◦C for later use. For the crystallization, the protein was mixed with 5 mM acarbose prior to setting up the screen. Initial screens were set up in a 96-well sitting drop format using commercially available screens. Initial hits were further refined in a 24-well hanging drop format. The best crystals grew in 0.1 M di-hydrogen phosphate, 1.8 M ammonium sulphate. Crystals were cryoprotected by addition of ethylene glycol to a final concentration of 15%. The crystals were flash frozen in liquid nitrogen prior to data collection.

#### 4.3.3. CfAM

Prior to crystallization, the protein was concentrated to 22.5 mg/mL by ultrafiltration in an Amicon centrifugation filter unit (Millipore), aliquoted to 50 μL; aliquots that were not immediately set up for crystallization were flash frozen in liquid nitrogen and stored at −80◦C to use later in optimizations. Initial crystallization experiments were carried out in the presence or absence of 4 mM CaCl2 and 40 mM acarbose. An initial hit was obtained for an acarbose complex, in just one condition (H3, Bis-tris 5.5, 25% w/v PEG3350) of JCSG screen (Figure 6a), out of total 192 conditions in two initial screens set up – JCSG and PACT premier™ HT-96 (Molecular Dimensions (Suffolk, UK)). The crystals were imperfect and were used to make the seeding stock. The seeding stock was prepared and microseed matrix screening (MMS, recent review in [14]) carried out using an Oryx robot (Douglas Instruments (Hungerford, UK)) according to the published protocols [36,37]. Briefly, crystals were crushed, and diluted with ~50 μL of mother liquor. The solution was transferred into a seed bead containing reaction tube and vortexed for three minutes. The seeding stock was used straightaway, and the remaining seeds were frozen and kept at ™20 ◦C. MMS was carried out in the PACT screen, giving a significant number of hits (Figure 6b). Crystals from condition A11 were used to make a seeding stock for the next seeding round. This time it was not a "classical" MMS-seeding into a random screen, but rather seeding into an optimization screen based on the initial conditions, but with different pH, salts and PEGs/PEG concentrations. The crystallization drops contained 150 nl protein + 50 nl seeding stock + 100 nl mother liquor from a new random screen. The final, good quality crystal was obtained in 12% PEG 3350 0.2 M NaNO3, CAPS pH 11.0 (Figure 6c).

**Figure 6.** Crystal optimization using microseed matrix screening.

#### *4.4. Data Collection and Processing*

The data were collected at Diamond on beam line I02, processed by XDS [38], and scaled with Aimless [39]. The statistics are shown in Table 1.


**Table 1.** Data collection and processing statistics.

Values for the outer shell are given in parentheses.

#### *4.5. Structure Solution and Refinement*

The structure of RpAM was solved by molecular replacement with Molrep [40], using TAKA amylase as template (PDB-ID:7taa). The structure of TeAM was solved with Molrep using the model of RpAM. The CfAM structure was solved using Molrep [40] with 3vx0 α-amylase from *Aspergillus oryzae* as a model. The final models were built using automated chain tracing with Buccaneer [41], followed by manual building in Coot [42], iterated with reciprocal space refinement using Refmac [43]. The statistics are summarized in Table 2.


**Table 2.** Structure solution and refinement.

Values for the outer shell are given in parentheses.

#### **5. Conclusions**

Taken together, we describe the structural and functional characterization of three novel fungal α- amylases with enhanced stability, of which two, CfAM and RpAM, have a higher pH optimum and greater temperature tolerance, well suited for usage in the detergent or saccharification industry. The structures reveal that these amylases follow the canonical domain structure of α-amylases, and that three shortened loops between β2/α<sup>3</sup> and in subdomain B are likely to be responsible for the altered enzymatic properties of the amylases compared to TAKA-amylase. For the first time, we have unambiguously identified up to three different N-glycosylation sites in α-amylases in the structures. Furthermore, the observed formation of an isoaspartate from an asparagine in one of the shortened loops might play a functional role. The complexes with acarbose derived transglycosylation products define seven subsites of the substrate binding crevice and helped to identify the catalytic residues unambiguously. In addition, a new previously unobserved carbohydrate binding site was revealed in the C-terminal β-sandwich domain of CfAM, which might be important for the initial interaction with its polymeric substrate.

#### **6. Patents**

The *Rhizomucor pusillus* amylase and the use of this amylase in various industrial applications have been claimed in patent application WO2006065579. A close homologue of the *Thamnidium elegans* amylase was claimed in patent application WO2006069290 including the use in industrial applications. **Author Contributions:** C.R. and O.V.M. analyzed the data, built and refined the structure and prepared the original draft. J.P.T. collected and analyzed the X-ray data. O.V.M., E.B., A.A. and J.W. crystallized the amylases and solved the initial structures. L.M. and S.T. cloned, produced, purified and characterized the amylases biochemically, C.R., C.A., G.J.D. and K.S.W. wrote, analyzed, and reviewed all stages of the manuscript. G.J.D. C.A. and K.S.W. planned and supervised the work.

**Funding:** This research received no external funding.

**Acknowledgments:** The authors are grateful for financial support by Novozymes. We thank ESRF for the access to beamline ID29 and Diamond Light Source for access to beamline I02 (proposal numbers mx-1221 and mx-9948) that contributed to the results presented here. The authors also thank Sam Hart for assistance during data collection.

**Conflicts of Interest:** The authors declare no conflict of interest, but we note that the Novozymes authors declare the following competing financial interest(s): Novozymes are a commercial enzyme supplier. Novozymes, provided the enzyme samples used for crystallization, did the functional characterization and provided the financial support for the project.
