Optimization of oxalic acid production by fungi for biotechnological solubilization of rock phosphate

ABSTRACT: This study describes a biotechnological strategy for producing and applying oxalic acid to solubilize phosphorus (P) from rock phosphate (RP). We evaluated six fungal species (Aspergillus niger FS1, Penicillium islandicum FS41, Pleurotus ostreatus PO1, Rhizoctonia solani Rhiz555, Sclerotium rolfsii Sr25, and Sclerotinia sclerotiorum Scl134) and three culture media (potato dextrose broth, Tsao and Strasser media) to maximize oxalic acid production. Among the fungal isolates tested and culture media, S. rolfsii Sr25 and Tsao medium showed efficient oxalic acid production. Tsao medium was optimized following a response surface methodology after initial screening of factors affecting RP solubilization. The optimized concentrations were 1 g L–1 NaNO3, 100 g L–1 glucose, 2 g L–1 KH2PO4, 4.5 g L–1 yeast extract, and 25 mg L–1 MgSO4·7H2O used for 20 days of incubation. Under these conditions, 71 mmol L–1 oxalic acid was obtained, representing a three-fold increase over production under non-optimized conditions (20 mmol L–1). Under optimized conditions, oxalic acid produced by S. rolfsii Sr25 reacted with low-solubility RP and solubilized 100 % of the P contained in ore. Thus, using S. rolfsii Sr25 to produce oxalic acid seems a promising biotechnological alternative for P solubilization from RP.


Introduction
Adequate phosphorus (P) supply is vital at all stages of plant development (Grant et al., 2001).However, P forms that are readily absorbable by plants are deficient in most soils due to P fixation on soil particles.This is aggravated in tropical soils, where phosphate is chemically adsorbed to aluminum (Al) and iron (Fe) oxides (Fontes and Weed, 1996).One of the predominant methods to overcome this deficiency is the use of soluble phosphate fertilizers.On tropical soils, these fertilizers are typically applied at dosages higher than the requirement of crops as a large part of the applied phosphate is adsorbed by soil components (Roy et al., 2016).
The primary P source for the production of fertilizers is rock phosphate (RP), a non-renewable resource that is costly and has nevertheless been increasingly exploited (Cordell et al., 2009).New technologies to reduce RP processing costs may substantially affect fertilizer production and reduce costs for farmers by facilitating the use of low-quality RP reserves, which are currently neglected due to technical difficulties and lack of profitability for its exploration (Mendes et al., 2020(Mendes et al., , 2021b)).
Solubilization of RP by microorganisms is a promising method to produce phosphate fertilizers using renewable carbon sources and low-cost substrates (Mendes et al., 2020(Mendes et al., , 2021b)).In these biotechnological systems, cultivation conditions can be modified to increase the synthesis of organic acids by microorganisms, which can be used for the solubilization of phosphate minerals.Among microbial organic acids, oxalic acid shows high acidity to RP solubilization (pK a1 1.25, pK a2 3.81) (Lide, 2004) with the potential for complexing Ca 2+ in the apatite mineral that forms the rock (Kpomblekou-A and Tabatabai, 1994;Mendes et al., 2021a, b).
Several fungi produce oxalic acid, by including ectomycorrhizal fungi, wood-degrading basidiomycetes, saprophytes, and phytopathogens (Dutton and Evans, 1996).Thus, oxalic acid-producing fungi show great biotechnological potential for phosphate solubilization.This research evaluated a high-performance biotechnological method of producing oxalic acid using fungi for RP solubilization.

Microorganisms
Isolates of Rhizoctonia solani (Rhiz555), Sclerotium rolfsii (Sr25), and Sclerotinia sclerotiorum (Scl134) were obtained from the Embrapa (Brazil) collection.Aspergillus niger (FS1) and Penicillium islandicum (FS41) were obtained from the collection of the Laboratory of Microbial Ecology, Federal University of Viçosa.Pleurotus ostreatus (PO1) was obtained from the collection of the Laboratory of Microbiology and Phytopathology, Federal University of Uberlândia.These fungal species have been reported as good oxalic acid producers (Durman et al., 2005;Mendes et al., 2014;Punja and Jenkins, 1984;Strasser et al., 1994;Tsao, 1963;Yang et al., 1993).The isolates were kept at 28 °C on Petri dishes with a culture medium of potato dextrose agar (PDA).

Culture media and cultivation conditions
The following culture media previously reported as stimulating fungal oxalic acid production were examined: a) potato dextrose broth with adequate conditions for oxalic acid production by different fungi, mainly of the genera Rhizoctonia (Yang et al., 1993) and Sclerotinia (Durman et al., 2005); b) Tsao medium (2 g NaNO 3 ; 1.5 g KH 2 PO 4 ; 1 g MgSO 4 •7H 2 O; 40 g glucose; 2 g yeast extract; 1 L distilled water), which elicits high production of oxalic acid by the fungus Pleurotus ostreatus (Tsao, 1963); c) Strasser medium (100 g sucrose; 1.5 g NaNO 3 ; 0.5 g KH 2 PO 4 ; 0.025 g MgSO 4 •7H 2 O; 0.025 g KCl; 1.6 g yeast extract; 1 L distilled water), as optimized for oxalic acid production by A. niger (Strasser et al., 1994).Tsao medium was modified for the optimization experiments as indicated in the respective sections.Culture media were placed in 125-mL Erlenmeyer flasks and autoclaved at 121 °C for 20 min.

Selection of fungi and culture media
The best fungus and culture medium combination for oxalic acid production was screened in 125-mL Erlenmeyer flasks containing 30 mL of PDB, Tsao, or Strasser media.Flasks were inoculated with three mycelial plugs (5 mm) of each fungus taken from the edges of colonies grown in PDA for 7 days at 28 °C.Inoculated flasks were incubated for 15 days in an orbital shaker at 240 rpm and 28 °C.Uninoculated flasks were used as controls.
The experiment was set up in a completely randomized design in a 6 × 3 factorial scheme with three replications.Data were subjected to the analysis of variance (ANOVA) and treatments were compared using the Fisher's least significant difference (LSD) test (p < 0.05).

Optimization of oxalic acid production by S. rolfsii Sr25
The combination of fungus (S. rolfsii Sr25) and culture medium (Tsao) that produced the highest amount of oxalic acid was selected for optimization.Initially, the effect of each component of the Tsao culture medium, as well as the effect of Mn addition (Pedersen et al., 2000;Ruijter et al., 1999), pH buffering (Strasser et al., 1994), and incubation time on oxalic acid production by S. rolfsii Sr25 were tested in a two-level fractional factorial experiment (2 8-4 , 1/16 fraction, resolution 4) with two replications of the complete experiment, totaling 32 runs.Each factor was assessed at two levels, coded as -1 (lower level) and +1 (upper level) in the experimental design (Table 1).
The experiment was conducted using 125-mL Erlenmeyer flasks containing 30 mL of culture medium.To produce a pH-buffered medium, components were diluted in 0.1 mol L -1 Tris-HCl buffer, whereas the unbuffered medium was prepared using distilled water.
In both cases, the initial pH was adjusted to 7.0.The medium was autoclaved (121 °C, 20 min) and inoculated with three mycelial plugs (5 mm) of S. rolfsii Sr25 taken from the edges of colonies grown in PDA for 7 days at 28 °C.
The experimental design and analysis of the effect of each variable on oxalic acid biosynthesis were performed using the Design of Experiments (DOE) procedure in Minitab 18 software.The main effect of each factor was calculated according to Eq. (1) (Almeida et al., 2014;Mendes et al., 2015): where: E m is the main effect, m + is the average amount of oxalic acid produced at the upper level of a factor, and m -is the average amount of oxalic acid produced at the lower factor level.The effect values calculated were subjected to the t test to evaluate whether they differed from zero at a 5 % probability level.
Factors that affected oxalic acid biosynthesis were selected for optimization by the response surface methodology (RSM).The factor levels in the design were calculated based on the factor increment with the highest product between the variation unit and the estimated regression coefficient obtained in the screening experiment (Table 2) (Box and Wilson, 1951).In this calculation, the central point (coded as 0) of a central composite design (CCD) (Myers et al., 2016) was defined as the highest or lowest factor level in the screening experiment when its effect on oxalic acid biosynthesis was positive or negative, respectively.The CCD was completed with two levels above the central point (coded as 1 and 2) and two levels below (coded as -1 and -2) (Table 3), which were determined according to the relative increments calculated (Table 2).Factors with no effect were added to the most economical conditions, that is, NaNO 3 at the lowest concentration (1 g L -1 ) and initial pH 7 without buffering.The experiment was conducted using 125-mL Erlenmeyer flasks containing 30 mL culture medium, according to CCD combinations (Table 3).The media were inoculated and incubated as described for the screening experiment.Data of oxalic acid biosynthesis by S. rolfsii Sr25 were used to adjust a polynomial quadratic model with a multiple regression procedure using the least squares method.The largest model was as described in Eq. ( 2) (Almeida et al., 2014): where: Ŷ is the estimated concentration of oxalic acid produced by S. rolfsii Sr25, i and j assume a value between one and the total number of factors (n), β 0 is the intercept, β i is the linear coefficient, β ii is the coefficient of the quadratic term, β ij is the interaction coefficient, x i and x j are the factor levels, and ε ij is the experimental error.An ANOVA of the adjusted coefficients and lack-offit tests was performed on the adjusted model (p < 0.05).Coefficients with a p-value greater than or equal to 0.1 were removed from the model.The experimental design and all calculations and analyses were executed in the DOE option of Minitab 18.The calculations were done according to Box and Wilson (1951).a Related to the experiment shown in Table 1, representing the average value between the levels -1 and +1.b Change in oxalic acid biosynthesis per unit, obtained in the screening experiment.c Changes in factor levels were calculated based on the change in the level of the factor with the highest regression coefficient in the screening experiment, that is, incubation time.

RP solubilization using fungal oxalic acid
The combination of factor levels that resulted in the highest production of oxalic acid (71 mmol L -1 ) by S. rolfsii Sr25 in the response surface experiment was used in the RP solubilization experiment: 1 g L -1 NaNO 3 , 100 g L -1 glucose, 2 g L -1 KH 2 PO 4 , 4.5 g L -1 yeast extract, and 25 mg L -1 MgSO 4 •7H 2 O.The experiment was carried out using 125-mL conical flasks with 30 mL of sterile culture medium and with the pH adjusted to 7, but without buffering.Flasks were inoculated and incubated as described for the screening experiment.
Rock phosphate was added to the medium at a concentration of 10 g L -1 at the beginning of the incubation together with the fungus (one step) or 20 days after inoculation of the medium with the fungus (two steps).RP was sterilized together with the culture medium in the one-step system, while in the two-step system, RP was sterilized separately by autoclaving at 121 °C for 20 min.Uninoculated flasks containing a sterile mixture of RP and liquid medium were incubated as controls.The RP sample was obtained in the municipality of Pratápolis, Minas Gerais State, Brazil (20°47'58.2"S, 46°50'39.9"W, altitude 836 m).It contained 10.5 % P, with particles measuring < 63 µm.The experiment was conducted in five replications per treatment.Differences between treatments were tested using the ANOVA, followed by the Tukey's post hoc test (p < 0.05) using the Experimental Designs package in R version 3.6.3.

Analytical methods
At the end of the incubation period in the oxalic acid production experiments, all samples were vacuumfiltered through a 0.2-µm membrane.Fungal biomass retained in the filter was transferred to a crucible and dried to constant weight in an oven at 80 °C.The filtrate was analyzed to quantify the amount of produced oxalic acid and measure the pH.Oxalic acid was quantified using capillary electrophoresis.Electropherograms were produced using capillary electrophoresis equipment with two detectors of capacitively coupled contactless C 4 D conductivity (Francisco and Lago, 2009;Silva and Lago, 1998), which were positioned near the capillary at 10 cm from each end.The fused silica capillary used in all experiments was 50 cm long and 50 µm i.d.× 375 µm o.d.The effective lengths were 10 and 40 cm for the first and second detectors, respectively.Before use, the capillary was washed using deionized water for 10 min, 0.1 mol L -1 NaOH for 10 min, deionized water for 10 min, and background electrolyte (running buffer) for 10 min.
At the end of the 30 day of the RP solubilization experiment, all samples were filtered using quantitative filter paper (25 µm pore size) for the pH and P analysis.Soluble P was determined by molecular absorption spectrophotometry using the ascorbic acid method (Braga and Defelipo, 1974).

Selection of fungi and culture media for oxalic acid production
The highest oxalic acid production (20 mmol L -1 ) was achieved by S. rolfsii Sr25 in Tsao medium (Figure 1A).Similar oxalic acid concentrations were produced by A. niger FS1 and S. rolfsii Sr25 in PDB medium.In general, fungi that produced more oxalic acid induced more significant acidification of the medium (Figure 1B).The Strasser medium stimulated biomass production; however, without increasing oxalic acid production (Figure 1C).

Optimization of oxalic acid production by S. rolfsii Sr25
Only five evaluated factors affected oxalic acid production by S. rolfsii Sr25 (Table 1).Higher dosages of glucose, KH 2 PO 4 , and yeast extract and longer incubation time increased oxalic acid biosynthesis, whereas a higher dosage of MgSO 4 •7H 2 O decreased the production (Figure 2).The other factors (pH, NaNO 3 , and MnCl 2 •4H 2 O) did not influence oxalic acid production.
Three-fold higher concentrations of oxalic acid (an increase from 20 to 71 mmol L -1 ) were produced using RSM to optimize the levels of factors affecting oxalic acid biosynthesis (Table 3).Contour plots show the optimal ranges of each factor (Figure 3).The highest oxalic acid production was observed at approximately 2.0 g L -1 KH 2 PO 4 and after approximately 20 days of incubation.The optimal concentration of yeast extract ranged from 4.0 to 4.5 g L -1 and MgSO 4 •7H 2 O was required at concentrations below 25 mg L -1 .

RP Solubilization using fungal oxalic acid
No difference between the one-step and two-step RP solubilization processes was observed (p > 0.05), with a mean concentration of 1,832 mg L -1 soluble P (Figure 4).On average, 1,395 mg P L -1 was solubilized by S. rolfsii Sr25 by subtracting the soluble P added to the medium as KH 2 PO 4 .The amount of RP added yielded 1,050 mg L -1 P, indicating that S. rolfsii Sr25 solubilized 100 % of the P present in RP.
The pH values of the culture medium differed (p < 0.05) between treatments at the end of the incubation period, with values of 2.4, 3.4, and 5.3 for the one-step, two-step, and control treatments, respectively.

Discussion
Solubilization of 100 % of the P in low-solubility RP was achieved using oxalic acid produced by S. rolfsii Sr25.An initial screening showed that oxalic acid production by S. rolfsii Sr25 was affected by the concentrations of glucose, KH 2 PO 4 , yeast extract, and MgSO 4 •7H 2 O and by incubation time.The use of RSM tripled oxalic acid production by S. rolfsii Sr25 from 20 to 71 mmol L -1 .The highest oxalic acid biosynthesis was achieved using 1 g L -1 NaNO 3 ; 100 g L -1 glucose; 2 g L -1 KH 2 PO 4 ; 4.5 g L -1 yeast extract; 25 mg L -1 MgSO 4 •7H 2 O, and incubation for 20 days.
Oxalic acid has excelent potential to solubilize RP of different origins and reactivities (Mendes et al., 2020).Moreover, oxalic acid was more efficient than sulfuric acid, the traditional reagent used for RP solubilization to produce fertilizers, releasing more P per mmol acid (Mendes et al., 2020).The reaction between oxalic acid and apatite reached complete P extraction at stoichiometric proportions of reagents.In addition, the reaction was fast, and most P was solubilized in the first hours, releasing up to 75 % of P after 1 h and 88 % after 24 h (Mendes et al., 2021b).These results support our findings, as we observed that 100 % of the RP were solubilized using oxalic acid produced by S. rolfsii Sr25.Medium acidification is one of the mechanisms of apatite solubilization in RP, as well as the capacity of oxalate to form a sparingly-soluble complex with calcium (CaC 2 O 4 ) (Kpomblekou-A and Tabatabai, 1994;Mendes et al., 2020Mendes et al., , 2021a, b), b).These features show the efficiency of oxalic acid in P solubilization, as evidenced by our results.Moreover, none of the organic acids commonly associated with phosphate solubilization (citric, gluconic, and malic acids) were detected in the strain of S. rolfsii Sr25 used in our study (data not shown), which suggests that oxalic acid was the primary fungal metabolite involved in RP solubilization.The amount of soluble P measured in the solubilization experiment was higher than the amounts of P added in the form of KH 2 PO 4 and RP (Figure 4).Possibly, this excess of P originates from the mineralization of organic P in yeast extract in the culture medium (Hidayat et al., 2006;Thompson et al., 2017).
Under optimized conditions, oxalic acid production by S. rolfsii Sr25 was tripled, reaching a concentration of 71 mmol L -1 and the crucial optimization factors were concentrations of KH 2 PO 4 , glucose, yeast extract, and MgSO 4 •7H 2 O and incubation time.NaNO 3 induces expression of the oah gene, which encodes oxaloacetate  hydrolase (Pedersen et al., 2000).However, in our study, an increase in NaNO 3 concentrations above 1 g L -1 did not result in higher oxalic acid production by S. rolfsii Sr25.Thus, NaNO 3 concentrations lower than in Tsao medium  appear to supply enough N. Oxalic acid production increased when more significant amounts of P were used, similar to reports for A. niger, where the accumulation of oxalic acid occurred in the presence of 2.5 g L -1 KH 2 PO 4 (Kubicek et al., 1988).In our study, only a small amount of Mg was necessary, or may not have been required, for oxalic acid biosynthesis by S. rolfsii Sr25.Production of organic acids by A. niger is stimulated in the presence of C excess (Papagianni, 2007;Strasser et al., 1994).Our results demonstrate a similar mechanism in S. rolfsii Sr25.At higher dosages of glucose and yeast extract than in the Tsao medium (2.5-and 2.25-fold increase, respectively), the biosynthesis of oxalic acid tripled.
In general, optimal oxalic acid production by fungi occurs between 10-15 days of incubation (Amadioha, 1993;Tsao, 1963).In our study, the optimal duration of S. rolfsii Sr25 incubation for oxalic acid production was approximately 20 days.Moreover, the culture medium containing the fungus assumed gel consistency after 20 days due to scleroglucan production (Castillo et al., 2015;Fariña et al., 1998;Survase et al., 2007), which is an undesirable characteristic regarding the subsequent application of oxalic acid-rich medium for RP solubilization due to the difficulty of homogenization.
Manganese stimulates oxaloacetase activity in A. niger (Pedersen et al., 2000;Ruijter et al., 1999); however, in our study, Mn addition did not affect oxalic acid production by S. rolfsii Sr25.Furthermore, contrary to observations on A. niger (Strasser et al., 1994), the pH buffering close to neutral values did not increase oxalic acid production by S. rolfsii Sr25.The optimum pH range for oxalate production was proposed at 5 to 8 (Ruijter et al., 1999), with the highest oxalate accumulation at approximately pH 6 (Kubicek et al., 1988).No corresponding effect was observed in our study.This discrepancy may be because oxalic acid is a product of the glyoxylate pathway of the tricarboxylic cycle in S. rolfsii and glyoxylate dehydrogenase is responsible for oxalic acid production (Maxwell and Bateman, 1968), whereas in A. niger oxalic acid is produced by oxaloacetase (Kobayashi et al., 2014;Kubicek et al., 1988).
Our findings indicate that oxalic acid may be an effective biotechnological option for solubilizing P from RP, potentially replacing sulfuric acid (Mendes et al., 2020).The oxalic acid concentration required to attain 100 % RP solubilization (71 mmol L -1 ) was achieved by S. rolfsii Sr25 in a culture medium using renewable carbohydrates, such as glucose and yeast extract.Although S. rolfsii Sr25 is phytopathogenic and can cause considerable losses to agricultural production, its use in a controlled production system may help explore its physiology for the benefit of agriculture.This study provides insights to support oxalic acid production by S. rolfsii Sr25 in a two-stage bioprocessing scheme aiming at producing P fertilizer.In the first stage, oxalic acid was produced by the fungus using renewable organic substrates as energy and carbon sources.In the second stage, the medium containing oxalic reacted with RP to generate a product with soluble phosphate that could be applied as a liquid fertilizer (Mendes et al., 2017) or dried in a subsequent step to produce solid fertilizer formulations (Gilmour, 2013).In this bioprocessing scheme, fungal biomass could be removed by filtration after the first stage to prevent fungal propagules from being disseminated with the fertilizer.
Currently, industrial production of oxalic acid largely relies on the oxidation of carbohydrates using sulfuric and nitric acids (Riemenschneider and Tanifuji, 2011).Furthermore, industrial RP solubilization for fertilizer production is carried out with sulfuric acid, produced from elemental sulfur mainly from the oil industry (Gilmour, 2013).Thus, the biotechnological scheme proposed here for RP solubilization using oxalic acid of microbial origin could eliminate the need for sulfuric acid in both industrial processes and represent a renewable and ecological alternative for producing phosphate fertilizers.

Figure 1 -
Figure 1-Oxalic acid biosynthesis (A), the pH of the culture media (B), and fungal biomass (C) produced by fungi in different culture media.Lowercase letters compare different fungi within each culture medium, while uppercase letters compare culture media for each fungus (Fisher's LSD test, p < 0.05).

Figure 2 -
Figure 2 -Oxalic acid biosynthesis by Sclerotium rolfsii Sr25 as affected by the concentration of glucose, NaNO 3 , KH 2 PO 4 , yeast extract, MgSO 4 .7H 2 O, and MnCl 2 .4H 2 O by the pH buffering (U = unbuffered, B = buffered) and incubation time.The dashed line represents the mean biosynthesis of oxalic acid.

Figure 4 -
Figure 4 -Soluble P in culture medium after incubation of rock phosphate (RP) with oxalic acid produced by Sclerotium rolfsii Sr25.In the one-step process, RP was added at the beginning of the incubation period, while in the 2-step process, RP was added on the 20 th day of incubation.The dashed line represents the dose of soluble P added as a component (KH 2 PO 4 ) of the culture medium for oxalic acid biosynthesis.Error bars represent standard deviation (n = 5).

Table 1 -
Screening of factors controlling the biosynthesis of oxalic acid by Sclerotium rolfsii Sr25.Significant by the t test (p < 0.05).a Effect = m + -m -(m + and m -represent the average oxalic acid produced for the upper and lower level of each factor, respectively). *

Table 3 -
Oxalic acid biosynthesis by Sclerotium rolfsii Sr25 under different combinations of factor levels in a central composite design (CCD).

Table 2 -
Calculation of factor levels for experimental design to fit response surface.The changes in factor levels were based on a 5-unit increment in time, which was the variable with the largest regression coefficient.Data used in the calculations were obtained in the screening experiment.