Conversion of Komagataella phaffii Biomass Waste to Yeast Extract Supplement

Response to Reviewer 1 Comments

1. Summary

Thank you for reviewing the manuscript and for identifying calculation and transcription errors in the originally submitted files. The author apologises for these mistakes, which have been corrected. All Bradford calculations have been reprocessed from the original cuvette absorbances using the recorded per-sample dilution factors and the reported standard curve (Abs = 2.1658·conc + 0.1197, R² = 0.9902). Corrected values appear in the revised manuscript (Table 3.4.1) and the audit materials are provided in Supplementary_Audit.zip. These numerical corrections do not alter the core conclusions of the work; the corrected calculations and provenance are provided to enable reproduction and verification.

2. Point-by-point response to Comments and Suggestions for Authors

Comment 1: 0.015% corresponds to 150 mg/L, not 15 mg/L; 15 mg/L equals 0.015 g/L and therefore 0.0015%. This discrepancy represents a 10-fold difference and significantly impacts the interpretation of the results and generates a lack of confidence in the data. Based on the amount of commercial product used (10 g/L), 150 mg/L might be more plausible. It is difficult to reconcile how 15 mg could result in similar performance to 10,000 mg.

Response 1: Thank you for noting the typographical error. All instances have been corrected to 15 mg·L⁻¹. The apparent paradox of a small recycled-YE dose producing similar culture performance to 10 g·L⁻¹ of commercial yeast extract is addressed by distinguishing bulk mass from bioavailability/composition. The Bradford total-protein assay and similar bulk metrics quantify total soluble protein but do not resolve the immediately bioavailable fraction (free amino acids, small peptides, vitamins, cofactors). The recycled yeast extract is derived from spent biomass of a primary K. phaffii fermentation expressing the same product and therefore plausibly contains enriched soluble peptides, residual vitamins/trace metals, and possibly trace amounts of residual secreted product — constituents that can have disproportionate effects on subsequent cultures at low addition rates. The revised manuscript notes this as a hypothesis (not a demonstrated fact).

Comment 2: Line 71: Which previous study?

Response 2: The mentioned work is one currently under review with MDPI Fermentation. Mention of this work has been removed from Line 71.

Comment 3: Line 94 and Table 2.3: the text states 1 g/L, whereas the table states 10 g/L ("YPD 10 g/L recycled yeast extract"). Which value is correct? These value discrepancies (on the order of 10×) undermine confidence in the study.

Response 3: Thank you for noting this error. The values have been corrected to 10 g/L.

Comment 4: On which study was the protocol described in section 2.5 based?

Response 4: This protocol is based upon the work of Offei, et al (2022). This reference has now been included in section 2.5 (now section 2.4 following other revisions).

Comment 5: There are different forms of 4-NPG. The authors must specify which was used (I assume β). Moreover, the compound should be written out in full the first time it appears in the text.

Response 5: It is the β form used. The full compound name has been added to section 2.4.

Comment 6: Line 209: Media is already the plural of medium, so "medias" is not correct. A careful review of the entire manuscript for such issues would be beneficial.

Response 6: All mentions of “medias” has been corrected to “media”.

Comment 7: Line 226: Why were the authors restricted to only 7 hours? Looking at the graph in Figure 3.2.1, it is not clear why the doubling time in YPD-C would be longer

Response 7: Thank you for the request for clarification. The experiment was not restricted to 7 hours: the full time course was 144 hours. The reference to “seven hours” in the text describes the difference in doubling times between media, not the duration of the experiment. Doubling times were calculated from the exponential growth phase (as described in the Methods) and YPD-C exhibited the longest doubling time; the other media showed doubling times that were within seven hours of the YPD-C value.

                             The originally reported doubling times were obtained from GraphPad Prism using a Malthusian (exponential) fit applied to the entire 0–144 h time course (the approach used in the initial submission to avoid excluding any data). Inclusion of late timepoints that reflect deceleration toward carrying capacity reduces the fitted exponential rate constant and therefore increases the calculated doubling time. Komagataella phaffii grown in YPD-C reached substantially higher biomass early in the run (≈10 OD units higher at 48 h), so a larger fraction of the 0–144 h series for that condition lies in the decelerating/near-stationary region; fitting the full series therefore yields an apparently slower (longer) doubling time for YPD-C.

Comment 8: Figure 3.2.1: What are the units for each value in panel B? Do these values differ significantly in statistical terms? Furthermore, the calculations for obtaining these values need clarification (it is not sufficient to state that they were "calculated using Malthusian exponential growth analysis on GraphPad Prism").

Response 8: Thank you — the figure legend lacked necessary detail. Panel B reports the fitted parameters from the Malthusian (exponential) model and their units: Y0 is the model intercept (predicted initial response at t = 0) in the units of the response variable (OD₆₀₀), k is the exponential rate constant (units h⁻¹), and Doubling time is t_d = ln(2)/k expressed in hours. The parameters were obtained by fitting the exponential model N(t)=Y₀e^kt to the full 0–144 h time course in GraphPad Prism (Malthusian model). Figure 3.2.1 Panel B has been updated to include these units, and the figure legend has been expanded with more detail on the calculation.

                             Per the reviewer’s request, per-replicate exponential rate constants (k, h⁻¹) were extracted and compared across media. No statistically significant difference was detected (one-way ANOVA / Tukey’s post-hoc test , p > 0.05). Per-replicate exponential rate constants (k, h⁻¹; n = 3 per group) were compared by Brown–Forsythe robust one-way ANOVA to account for possible heteroscedasticity: F(3,8) = 1.043, p = 0.4247. This was also performed for doubling times and similarly no significant difference was found: F(3,8) = 0.9777, p = 0.4500.  Full per-replicate k values and per-replicate doubling times, along with the calculated p-values from the ANOVA are provided in the Supplementary Material file.

Comment 9: Table 3.4.1 shows that if PD medium contains 5.41 g/L of soluble protein and YPD-C contains 8.94 g/L, the difference (3.53 g/L) could be attributed to the 10 g/L of yeast extract in YPD-C. Following this logic, YPD+2%P would have the same 5.41 g/L of soluble protein from peptone and 3.23 g/L from only 15 mg/L of recycled yeast extract. Similarly, YPD+1%P would have 2.7 g/L from peptone and 3.91 g/L from only 15 mg/L of recycled yeast extract. These values are difficult to reconcile: how could 15 mg/L of any substance provide 3.23 g/L of soluble protein?

Response 9: The paradox was caused by two processing errors in the original spreadsheet: (i) an omitted per-sample dilution multiplier for YPD-C and (ii) transcription errors for several PD absorbances. All raw cuvette absorbances were reprocessed with correct dilution factors and the Bradford standard curve; corrected total-protein results are reported in revised Table 3.4.1: YPD-C = 26.84 ± 0.25 g·L⁻¹; YPD+1%P = 13.22 ± 0.18 g·L⁻¹; YPD+2%P = 17.28 ± 0.16 g·L⁻¹; PD = 15.44 ± 0.15 g·L⁻¹ (mean ± SD, n = 3). These values are consistent with the protein contribution expected from 2% (w/v) peptone (COA-derived crude protein ≈ 15.9 g·L⁻¹; Sigma-Aldrich Peptone No. 2, lot BCCC2601). Arithmetic check: 15 mg·L⁻¹ = 0.015 g·L⁻¹, so even if the recycled powder were 100% protein its maximum contribution would be 0.015 g·L⁻¹ — it cannot account for multi-g·L⁻¹ increases. The corrected calculation workbook and all raw data are provided in the supplementary files and audit zip. The greatly elevated protein in YPD-C is therefore attributable to the deliberate 10 g·L⁻¹ commercial yeast extract addition, plus routine assay/matrix effects. The discussion of the manuscript has been updated accordingly.

Comment 10: Line 359 and Table 3.4.2: Based on which previous study do the authors state that there are 3 g/L of reducing sugars from yeast extract and peptone? And even if that were the case, how do they explain that virtually all media contain ~23 g/L of sugar? It is particularly unclear how PD medium, with no yeast extract, could have the same reducing sugar concentration as YPD-C, which has the same peptone content plus 10 g/L of yeast extract. 

Response 10:  The sentence on line 359 that attributed an additional “~3 g·L⁻¹” reducing sugar to yeast extract/peptone was an unsupported assumption and has been removed. The revised manuscript now explains that the measured reducing-sugar concentrations are dominated by the intentionally added dextrose (20 g·L⁻¹ in the YPD recipe) and that the small excess (~23 g·L⁻¹ measured) is within the expected preparation and assay uncertainty for routine media preparation and reducing-sugar assays. Because the same dextrose stock and dispensing procedure were used for all media, similar measured sugar concentrations across PD, YPD-C and the recycled formulations are expected. Yeast extract and peptone contribute negligible reducing sugars to the media composition as they are primarily protein/peptide and nucleotide sources and contain only a modest carbohydrate fraction (mostly polysaccharides and bound sugars) [1, 2]. The discussion (lines 503 to 512) has been updated accordingly.

Comment 11: Lines 479–481: If the two media indeed have the same peptone concentration, it seems highly unlikely that a medium with 15 mg/L of recycled yeast extract could provide the same soluble protein content as one with 10,000 mg/L of commercial yeast extract.

Response 11: Please see response 9.

3. Additional clarifications

The corrected protein numbers are reported in Table 3.4.1 and corresponding audit materials are provided in the Supplementary archive. Although 15 mg·L⁻¹ cannot by mass balance explain a multi-gram increase in measured protein, it is plausible that a small addition of recycled yeast extract (sourced from spent biomass of a primary K. phaffii fermentation producing the same product) may exert a disproportionate biological effect because of its composition. Spent yeast/autolysates concentrate soluble peptides, free amino acids, nucleotides, B-vitamins and other small cofactors [1]. and recombinant K. phaffii cultures can release and retain extracellular protein in the spent broth [3], so spent media from a primary fermentation expressing the same enzyme may contain trace amounts of the target product and proteinaceous cofactors that can beneficially influence a secondary expression culture even when added at low mass. These compositional features and possible trace residual target protein can change bioavailability at low addition rates and are therefore offered here as a plausible hypothesis, and this has been reflected in the discussion section of the manuscript.

Response to Reviewer 3 Comments

1. Summary

Thank you very much for taking the time to review this manuscript. Please find the detailed responses below and the corresponding revisions/corrections highlighted in the re-submitted files.

2. Point-by-point response to Comments and Suggestions for Authors

Comments 1: Combine Sections 2.1 and 2.2 to improve the flow and avoid redundancy.

Response 1: Thank you. These sections have now been combined.

Comments 2: Shuffle Sections 2.6 and 2.7 to improve logical progression

Response 2: These sections have been swapped

Comments 3: Highlight the specific advantages of this method over existing recycling approaches beyond being described as “simple” and “reproducible.” For example, is it faster, more scalable, more cost-effective, or better suited for certain yeast strains?

Response 3: The advantages of this method has been more clearly stated in the discussion section lines 542 – 551.

Comments 4: Consider potential limitations such as High freeze-drying energy costs at large scale, Variability in nutrient profiles between fermentation batches and Restricted applicability beyond K. phaffii.

Response 4: The authors believe the limitations have been adequately covered in section 4.4 of the discussion.

Comment 5: The method has only been validated with K. phaffii; testing with other yeast or microbial species would strengthen the applicability of the findings.

Response 5: The reviewer is correct that demonstrating the approach in additional yeast or microbial hosts would strengthen the generalisability of the findings. The present study was intentionally designed as a focused proof-of-concept in Komagataella phaffii to demonstrate feasibility within that industrially relevant chassis. Extending experimental validation to other hosts would require host-specific optimisation (media, expression conditions and downstream handling) and resources beyond the scope of this manuscript. The revised Discussion now explicitly notes this limitation and frames cross-species validation as an important avenue for future work.

Comment 6: Include data on the shelf life and stability of the freeze-dried extract under different storage conditions to assess its practical utility

Response 6: The recycled yeast extract used in this study was stored frozen (−20 °C) for the duration of the experimental work; when thawed and left unfrozen the material loses its handling properties and becomes a fluffy, non-usable mush. No systematic shelf-life or accelerated stability testing was performed (these analyses are outside the scope and resources of the current revision). To address the reviewer’s concern the revised manuscript now documents the storage/handling procedure and our direct observation that the material must be maintained frozen for reliable use in methods section 2.1. Frozen storage at −20 °C strongly reduces chemical and microbial degradation and for many proteinaceous powders preserves functionality for months to years, but actual shelf-life depends on formulation, residual moisture and packaging and would therefore need to be demonstrated empirically for this specific recycled extract.

1. Summary

We are grateful for the very helpful review comments that have greatly improved the manuscript. Please find the detailed responses below and the corresponding revisions/corrections highlighted in the re-submitted files

3. Point-by-point response to Comments and Suggestions for Authors

Comment 1: The manuscript lacks essential analyses, including key conversion factors such as growth rate and cell yield per carbon source, as well as the monitoring of primary metabolites and hexose concentrations throughout cultivation. For a thorough evaluation of the feasibility of replacing commercial YPD with recycled spent yeast, further investigation of the freeze-drying process is necessary—particularly the quantification of amino acids, salts, and total protein content.

Response 1: We thank the reviewer for this insight and suggested new data acquisition. The growth rates (doubling time, k value) of all cultures have now been included as indicated by Panel B of figure 3.1. Cell yield per carbon source has been performed based upon the reducing sugar content of the media, along with the WCW (g) information already in the manuscript. This can be seen in Table 3.4.3

      The Lever assay (PAHBAH (p‑hydroxybenzoic acid hydrazide assay) was carried out to measure the reducing sugars present in the media types at T0 and T144. A Bradford assay was performed to indicate total protein content of each media type. And as mentioned, specific growth rates, doubling time, and cell-yield per carbon source have all been calculated and included in the manuscript. These tests give an indication of the overall composition of the medias. The protocols have been described in methods section 2.7 (Lines 128 – 151) and the corresponding results can now be seen in section 3.4 (Lines 324 – 392). These new assays show that all medias consume the same ~ 20 g/L glucose (PAHBAH), support equivalent protein uptake (Bradford) and that YPD-2% peptone confers growth yields (Yₓ/S) matching or exceeding commercial YPD. Together with our growth‑rate data, this provides a comprehensive functional characterization of recycled extract media without further HPLC or LC/MS profiling.

            Detailed profiling of every amino acid, metabolite, or salt can be avoided because our combined data set—showing equivalent glucose consumption, total protein uptake, biomass yield per glucose, and matched growth kinetics—demonstrates functional parity with commercial YPD. Nutrient composition will inherently vary with the source biomass, but matching the overall “nutrient load” (sugar and protein input/output) across diverse batches ensures reproducible performance without exhaustive chemical analyses. In rich media, it is the total pool of bioavailable carbon, nitrogen, vitamins, and cofactors—not the exact ratio of each individual component—that drives yeast growth, and our assays capture that system‐level equivalence.

3. Point-by-point response to Comments and Suggestions for Authors

Comment 1: Italicize Komagataella phaffii in the title. Please check the entire text for correct formatting.

Response 1: Thank you for bringing this to our attention, we will ensure correct formatting in the manuscript.

Comment 2: Lines 47-50: This sentence is confusing. The authors mention a nutritional variation between different biomass sources but then claim that residual biomass can be efficient in circular systems based on the same product. Please clarify this point.

Response 2: Thank you for this feedback, this sentence has been re-written to give more clarity on the point being made (lines 47-54)

Comment 3: Section 2.2: I believe the authors forgot to specify the pressure employed in the equipment.

Response 3: This information has been added in.

Comment 4: Wouldn't it be expected that the freeze-drying process used by the authors would preserve cell integrity (even if the cells were likely dead due to prior freezing at -20°C)? How was it possible to solubilize the "fine powder" mentioned? Also, it is important to highlight that freeze-drying is often used to preserve viable cells. Can the authors ensure that the recycled cells were dead? Furthermore, the methodology does not mention whether the media were sterilized in an autoclave.

Response 4: We thank the reviewer for this insightful comment. Although freeze‑drying alone can preserve intact cells, our protocol includes extended freeze‑thaw stress in stationary phase (–20 °C followed by –80 °C) of high‑OD cultures. Ice‑crystal formation during these cycles disrupts yeast membranes and weakens cell walls, triggering autolytic release of intracellular contents upon subsequent sublimation. The resulting lyophilisate is therefore largely cell‑components and soluble cytosol, which dissolves readily in water [1].

            All medias were autoclaved at 121°C and kept sterile prior to culturing experiments. This, combined with the robust freezing, gives confidence that the recycled cells were indeed dead as this standard sterilization condition is universally accepted to kill both vegetative cells and most spores in laboratory practice

Comment 5: 5) Section 2.3: Please organize the different media compositions into a table. Line 85: Were only the media containing 15 mg/L of extract supplemented with peptone? Please, rewrite this sentence.

Response 5: Table 2.3 has now been added to the methods section. The mentioned sentence has also been rewritten for clarity.

Comment 6: Table 3.1: Please include the standard deviations for each triplicate measurement.

Response 6: The standard deviations and additionally the coefficient of variance have now been included in table 3.1, along with a mention of this data in the preceding text in lines 184-181

Comment 7: Please clarify in the manuscript: Was only the condition with 15 mg/L of recycled yeast extract used for the cell growth analyses?

Response 7: This has been clarified in the manuscript on line 208. Four media types were used in this experiment – Commercial YPD media, Recycled YPD media at 15 mg/L with 1% peptone, Recycled YPD media at 15 mg/L with 2% peptone, and media containing only 2% peptone and 2% dextrose. Any other formulation of recycled YPD media was turbid and the yeast was out of solution, and lower concentrations were feared to not have enough nutrient to support growth (and as outlined, it was desired to have the optical densities match commercial YPD media).

Comment 8: Please add a graphical legend to Figure 3.1 to facilitate visualization (this is better than relying on the figure title/caption description).

Response 8: This figure has been updated accordingly

Comment 9: Line 265: The usual value is 1%, which equals 10 (ten) g/L. Please double-check the entire text to ensure consistency in the units used

Response 9: Thank you for noting this. The manuscript has been updated accordingly

Comment 10: How do the authors explain the fact that a much lower concentration of recycled extract supports the same cell growth as 10 g/L of commercial extract? Did the authors conduct tests with 15 mg/L of commercial extract? This point must be thoroughly discussed in the text, as it does not seem reasonable that 15 mg/L would promote the same cell growth as 10 g/L.

Response 10: We did not evaluate 15 mg/L of commercial yeast extract, as this concentration lies well below standard usage. Commercial preparations are typically recommended at 3–50 g/L (0.3–1 % w/v) for routine microbiological media, and our laboratory and other researchers routinely employs 10 g/L in YPD. At 15 mg/L (0.0015 % w/v), one would expect essentially no nutritional effect from the commercial product.

            Commercial yeast extract (e.g., Sigma Cat. No. 113885) contains ≈ 65 % protein/peptide, with the remaining mass comprising ash, moisture, and non‑nutritive ballast. To supply ~ 6.5 g/L of bioavailable protein/peptides, one needs ~ 10 g/L of powder. Our recycled extract is produced by freeze‑thaw–induced autolysis and lyophilization of high‑OD K. phaffii biomass. This is an essentially pure, autolysed yeast biomass with most of its ash, moisture, and other inert material removed— thus it delivers a much higher concentration of nutrients per milligram than commercial powder in a readily bioavailable form.

Comment 11: Lines 342-345: It is not clear from the text whether the authors autoclaved the media before analyzing cell growth. The only mention of autoclaving appears on line 261, where the authors state that autoclaving caused precipitation. As I mentioned earlier, freeze-drying can preserve some viable cells, and using non-autoclaved media may have led to misinterpretation of the results (in this case, the use of recycled cell extracts could act as an additional inoculum). Did the authors check whether the freeze-dried material contained any viable cells?

Response 11: For clarification, all medias were prepared (those with commercial and recycled yeast extract) and were then autoclaved. This information has been added to the methods section on lines 85 and 86.

1. Summary

We are grateful for the very helpful review comments that have greatly improved the manuscript. Please find the detailed responses below and the corresponding revisions/corrections highlighted in the re-submitted files

3. Point-by-point response to Comments and Suggestions for Authors

Comment 1: Title. Please write Komagataella phaffii in italics

Response 1: Thank you for bringing this to our attention, we will ensure correct formatting in the manuscript

Comment 2: The Abstract should be enriched with numerical results to highlight the main findings of the work. Additionally, keywords are used to increase the visibility of work in scientific searches. Repeating general keywords already in the title and Abstract is not advisable. Please consider removing or substituting "Komagataella phaffii" and "secondary fermentation" from the keywords

Response 2: Specific details have been added to the abstract. Keywords have been updated also. Thank you for these suggestions.

Comment 3: The introduction is adequate. It contains relevant information, providing a background sufficient for the manuscript. The word "scalable" should be removed [Line 59], as the work does not include any experiments or data to support the scalability of the described process.

Response 3: Thank you. We have removed the mention of scalability as you have suggested

Comment 4: The materials and methods are clearly described. Please write "K. phaffii" in italics [Line 93] and review it throughout the manuscript

Response 4: Thank you for bringing this to our attention, we will ensure correct formatting in the manuscript.

Comment 5: The results section is well-written. Please capitalize the word "table" [Line 121]. When a specific table (or figure) is mentioned, as in this case, "Table 3.1," it is considered a proper noun and should be capitalized. It must be corrected throughout the manuscript. On the other hand, please write "144-h" instead of "144-hr" since the correct abbreviation for "hour" is "h" [Lines 146, 164].

Response 5: Thank you. These changes have been made as suggested.

Comment 6: Discussion. Why do the authors believe they simulate the YPD medium by simply matching its optical density? What is the basis for this? It is not even mentioned in the discussion. Is there any precedent for something similar in the literature? On the other hand, what do the authors assume happens to the potential toxic metabolites produced in the initial fermentation?

Response 6: Thank you for bringing attention to this point that requires clarification. We are simulating YPD medium through adding in commercial peptone and dextrose, but recycled yeast extract from waste biomass. The formulated medias were measured using optical density to ensure that the added recycled yeast extract was soluble in solution, and that it had a density similar to commercial YPD media.

Monod (1949) showed that μ (specific growth rate) and cell concentration are tightly coupled to substrate availability, which can be monitored via turbidity changes [1]. ISO 7027 codifies the measurement of turbidity as a means to estimate suspended‐solid concentration in very heterogeneous samples [2]. There is precedent in literature for a similar approach, where complex media supplements are often standardized by turbidity or OD rather than by individual component assays, because their exact composition fluctuates batch‑to‑batch [3]. Commercial YPD (10 g/L yeast extract + 20 g/L peptone) yields a characteristic OD₆₀₀ when dissolved—reflecting its total soluble/colloidal content. By dissolving the recycled yeast extract powder to the same OD₆₀₀ we are functionally matching the total nutritional “load”, which—in a rich medium—is what determines growth rate and yield more than the precise ratios of peptides vs. vitamins vs. minerals.

            New data added into results section 3.4 highlights that the nutritional profile of the commercial yeast extract and the YPD-2% peptone recycled extract medias are very similar. This confirms that YPD is simulated effectively in our 15mg addition of recycled yeast extract.