Introduction

The methylotrophic yeast Komagataella phaffii (formerly known as Pichia pastoris) is a widely used host for producing heterologous proteins, due to its efficient protein secretion system that simplifies protein recovery, its ability to perform post-translational modifications, and its ability to be grown to a high biomass concentration in low-cost defined media. Furthermore, it can yield high levels of extracellular heterologous proteins in controlled bioprocesses (Cereghino and Cregg 2000). Usually, an expression cassette harboring a promoter sequence, the heterologous coding sequence, and a transcriptional terminator sequence is integrated into the yeast genome to express the heterologous gene during cell growth or in the presence of an inductor.

The transcription of the heterologous gene is a crucial step in recombinant protein production. Therefore, strong and controllable promoters are usually preferred for the efficient production of recombinant proteins (Vogl and Glieder 2013). Since the promoter from the alcohol oxidase 1 gene (PAOX1) is a strong and tightly regulated methanol-inducible promoter, it is the most commonly used promoter to produce foreign proteins in the K. phaffii expression system (Cereghino and Cregg 2000). Nevertheless, when methanol is used as a carbon source and inductor, some safety and operational constraints must be considered in large-scale bioreactors (Mattanovich et al. 2014). In contrast, the promoter from the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene (PGAP) is a strong and constitutive promoter that has emerged as an alternative to PAOX1 to avoid the use of methanol (García-Ortega et al. 2019). Nevertheless, a strong promoter is not always the most favorable choice for heterologous protein production, especially when the correct folding of the heterologous protein and/or post-translational processing in the secretory pathway is the limiting step for obtaining high amounts of the extracellular product (Cereghino and Cregg 2000; Mattanovich et al. 2004; Hohenblum et al. 2004). The high level of heterologous gene expression can overwhelm the post-translational machinery of the cell, causing a significant proportion of the foreign protein to be misfolded, unprocessed, mislocalized, and degraded. Weak promoters can be used for expressing genes that produce toxic proteins to the host cells (Ruth et al. 2010) and in synthetic biology applications that commonly require the co-expression of multiple genes (Vogl et al. 2016; Jin et al. 2019). To date, the lowest strength K. phaffii promoter is reported to be from the YPT1 gene (PYPT1), which encodes for a GTPase involved in secretion (Sears et al. 1998). This promoter provides constitutive gene expression levels in glucose cultures that are 100-fold lower than those obtained from PGAP, though it is far less studied than other promoters.

On the other hand, protein production levels are determined mainly by the synergistic action of the promoter and the transcriptional terminator (Matsuyama 2019). The main function of terminators is the termination of the transcription process, but the 3′-untranslated region (3′UTR) contained in the terminator sequence also modulates the half-life of mRNA, and transcriptional and translational efficiencies, nuclear export, and the cellular localization of mRNA (Kuersten and Goodwin 2003; Mayr 2019). Therefore, besides the promoter, the transcriptional terminator is another regulatory element in the expression cassette that can modify the expression levels of the gene of interest.

In mining homemade RNA-seq data from a K. phaffii KM71 strain grown in methanol, glucose, or glycerol, we observed that the PAS_chr3_1033 gene (hereafter 1033 gene) showed almost constant and low transcript levels. Therefore, the promoter of the 1033 gene (P1033) should be weak and constitutive. The 1033 gene (also referred to as PP7435_CHR3-0135 gene for the K. phaffii CBS 7435 genome) codes for a hypothetical protein with 38% and 29% sequence identity with D-amino-acid oxidases from the yeast Scheffersomyces stipitis CBS 6054 (Genbank accession no. XP_001384459.2) and K. phaffii CBS 7435 (Genbank accession no. CCA39107.2), respectively.

To fully characterize the P1033 and the transcriptional terminator of the 1033 gene (T1033), we identified the P1033 and T1033 sequences, along with the 5′UTR sequence, the transcription start site (TSS), the 3’UTR sequence and A- and T-rich sequences downstream of the stop codon for the 1033 gene. The sequence of the TATA-like motif, a mammalian-type initiator (INR) element, and the potential K. phaffii transcription factors (TF) for the P1033 sequence were also predicted. Furthermore, we constructed two K. phaffii strains with two combinations of regulatory DNA elements (i.e., P1033-TAOX1 and P1033-T1033 pairs; TAOX1: transcriptional terminator of the AOX1 gene) in the expression cassette that harbors the FTEII gene coding for a beta-propeller phytase (Viader-Salvadó et al. 2010) as the reporter gene, which was functionally associated with the alpha-factor prepro-secretion signal coding sequence. We selected clones of each construct with a single copy of the expression cassette integrated into the yeast genome and evaluated the impact of the regulatory DNA element combinations on the transcript levels of the heterologous FTEII gene in cells grown in glucose or glycerol as the carbon source. Moreover, we analyzed the transcript levels of the endogenous 1033 and GAPDH genes as transcript-level references. The extracellular product/biomass yields (Yp/x) were also determined to correlate it with the transcript levels.

Materials and methods

Strains, plasmids, enzymes, media composition, and chemicals

The K. phaffii KM71 (his4) strain was purchased from Thermo Fisher Scientific (Waltham, MA, USA). Plasmid pUCIDT-AMP used for cloning was from Integrated DNA Technologies, Inc. (Coralville, IA, USA). Plasmid pGAHFTEII was previously constructed in our laboratory (Herrera-Estala et al. 2022). This vector harbors an expression cassette containing the PGAP sequence followed by the Saccharomyces cerevisiae alpha-factor prepro-secretion signal coding sequence (flanked by AatII and XhoI restriction sites), a nucleotide sequence encoding the beta-propeller phytase FTEII with K. phaffii-preferred codons (Viader-Salvadó et al. 2010), the transcriptional terminator from the AOX1 gene (TAOX1), and a functional copy of the histidinol dehydrogenase (HIS4) gene. Endo H glycosidase, and AatII, XhoI, and Bsu36I endonucleases were from New England Biolabs (Beverly, MA, USA). SalI endonuclease was from Clontech (Palo Alto, CA, USA). M-MLV reverse transcriptase, RQ1 RNase-free DNase, GoTaq DNA polymerase, and oligo(dT)15 primer were from Promega (Madison, WI, USA). Other primers and probes were from Integrated DNA Technologies, Inc. (Coralville, IA, USA); sequences are described in Supplementary Table S1. Yeast extract-peptone-dextrose (YPD), regeneration dextrose base (RDB) agar, and buffer minimal glycerol (BMG) were prepared according to the manual from the Pichia Expression Kit (Thermo Fisher Scientific, Waltham, MA, USA). BMGly and BMGlc media were similar to BMG medium but with 30 mM glycerol (0.28% [w/v]) or glucose (0.54% [w/v]) instead of 1% (w/v) glycerol, respectively. All chemicals used were from Sigma-Aldrich Co. (St. Louis, MO, USA) or Productos Químicos Monterrey (Monterrey, Nuevo León, Mexico).

Promoter and transcriptional terminator sequences of the 1033 gene

The sequence located between the coding sequences (inter-CDS region) of the 1033 gene and the adjacent upstream gene of the K. phaffii GS115 chromosome 3 sequence (i.e., NOB1 gene) was considered the P1033 sequence. The sequence surrounding the TSS was identified by analyzing the data for five homemade RNA-seq of a K. phaffii KM71 strain grown in methanol or glycerol, available at NCBI Sequence read archive (SRA) under the BioProject accession number PRJNA930494. The HISAT2 program (Kim et al. 2015) (Galaxy version 2.1.0) was used to map the RNA reads to the inter-CDS region of the 1033 gene and the upstream gene at the K. phaffii GS115 chromosome 3 sequence. The mapping results were visualized with the Integrative Genomics Viewer (Thorvaldsdóttir et al. 2013) to detect a subregion of the inter-CDS region without aligned RNA reads. The adjacent downstream and upstream nucleotides of the 3′ end of the subregion with non-aligned RNA-reads were considered as the sequence surrounding the dominant TSS and the beginning of the 5′UTR sequence of the 1033 gene. The sequence of the TATA-like motif and a mammalian-type INR element downstream of the TATA-like box were predicted using the program YAPP Eukaryotic Core Promoter Predictor (http://www.bioinformatics.org/yapp/cgi-bin/yapp.cgi), which was also used to predict the specific dominant TSS nucleotide through a synergistic combination of the TATA-like box and the INR element. Moreover, an in silico analysis for predicting the putative transcription factor-binding sites (TFBS) in the P1033 sequence was performed using Transcription Factor Affinity Prediction Web tools (Thomas-Chollier et al. 2011) with the fungi [transfac_2010.1 + jaspar] matrix and the yeast promoters (only for fungi matrices) model, together with the Benjamini–Hochberg multiple correction test. Only results with p-values less than 0.05 were considered. The UniProt database (https://www.uniprot.org/) was used to retrieve the S. cerevisiae protein sequences involved with the TFBS. The K. phaffii orthologs of these sequences were identified by the Best Reciprocal Blast Hit technique (Ward and Moreno-Hagelsieb 2014) and were considered potential K. phaffii TF for the P1033 sequence.

The sequence from the 3′ end of the 1033 CDS until 50 nucleotides downstream of the 3′UTR of the 1033 gene was considered as the T1033 sequence. The 3′URT sequence was determined by analyzing the data for five homemade RNA-seq (BioProject accession number PRJNA930494 at NCBI SRA) as described above for 5′UTR determination mapping the RNA reads to the inter-CDS region of the 1033 gene and the downstream gene at the K. phaffii GS115 chromosome 3 sequence. The beginning of the subregion with non-aligned RNA-reads was considered as the 3′UTR end of the 1033 gene. Moreover, the T1033 sequence was analyzed in silico to find A- and T-rich sequences described in the literature with a high number of occurrences in the sequences located downstream of the stop codon in yeast genes (van Helden et al. 2000).

Construction of K. phaffii KM71/P1033-FTEII-TAOX1 and KM71/P1033-FTEII-T1033 strains

A synthetic DNA sequence harboring an AatII restriction site, the P1033 sequence, the coding sequence for the S. cerevisiae alpha-factor prepro-secretion signal, based on K. phaffii-preferred codons (De Schutter et al. 2009) until the XhoI site, a NotI restriction site, 11 spacer nucleotides, the T1033 sequence, and the sequence from the 3’ end of the TAOX1 until the Bsu36I site of the vector pPIC9 was synthesized, cloned into the vector pUCIDT-AMP and sequenced by Integrated DNA Technologies, Inc. (Coralville, IA, USA) to generate the plasmid pUCIDTP1033T1033. The liberated fragment from AatII and XhoI digestion of the plasmid pUCIDTP1033T1033 was ligated into the vector pGAHFTEII, previously digested by the same endonucleases, to produce the expression vector pP1033-FTEII-TAOX1. Similarly, vector pP1033-FTEII-T1033 was constructed by NotI and Bsu36I digestion of the plasmid pUCIDTP1033T1033 and ligation into the vector pP1033-FTEII-TAOX1. The correct constructions of the two vectors were confirmed by PCR using PP1033 and FTE2 primers, and FTE1 and 3TH primers, directed to the P1033 and the FTEII gene or to the FTEII gene and a region of the linearized vector integrated into the yeast genome downstream of the transcriptional terminator sequence, respectively. All DNA manipulations were performed according to standardized methods (Green and Sambrook 2012).

K. phaffii KM71 cells were transformed with SalI-linearized pP1033-FTEII-TAOX1 or pP1033-FTEII-T1033 DNA by electroporation according to the manufacturer’s instructions (Thermo Fisher Scientific, Waltham, MA, USA). Transformants were plated onto RDB agar plates at 30 °C for subsequent screening by histidine prototrophy. The integration of the expression cassette at the HIS4 locus of the K. phaffii genome in randomly selected His+ colonies from the two transformations was verified by PCR using the PP1033 and FTE2 primers, and FTE1 and 3TH primers.

Transformant selection from KM71/P1033-FTEII-TAOX1 and KM71/P1033-FTEII-T1033 constructions

Thirty-two and 20 transformants of each strain (KM71/P1033-FTEII-TAOX1 and KM71/P1033-FTEII-T1033, respectively) were grown in 10 mL of YPD medium at 30 °C and 250 rpm for 12–14 h until reaching an optical density at 600 nm (OD600) of 8–10. The grown cells harvested by centrifugation (3000 g, 10 min, 4 °C) were used to inoculate 25 mL of BMG medium supplemented with 0.1% (w/v) CaCl2, to an initial OD600 of 1.4. Further incubation was performed for 24 h at 30 °C and 250 rpm, with the addition of 10% (w/v) glycerol to a final concentration of 1% (w/v) after 14 h of culture. The biomass concentration was estimated based on 1.0 OD600 unit corresponding to 0.23 g dry cell weight (DCW) per liter (Herrera-Estala et al. 2022). The cell-free culture medium from each culture was recuperated by centrifugation (3000 g, 10 min, 4 °C), and the protein concentration was determined by the Bradford protein assay, using bovine serum albumin as the standard. The extracellular protein/biomass yield was calculated as the ratio of extracellular protein concentration to biomass concentration.

The three clones (His+ transformants) of each strain that yielded the lowest extracellular protein/biomass yield were analyzed by quantitative PCR (qPCR) to determine the copy number of the integrated expression cassette in the yeast genome as described previously (Herrera-Estala et al. 2022). One transformant of each constructed strain with a single copy of the FTEII expression cassette was selected for further experiments.

P1033 and T1033 functionality analysis

The P1033 and T1033 functionalities on the heterologous gene were verified by detecting the FTEII transcripts using reverse transcription polymerase chain reaction (RT-PCR), and FTEII protein using SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and phytase activity.

Total RNA was isolated as described previously (Caballero-Pérez et al. 2021), from KM71/P1033-FTEII-TAOX1 and KM71/P1033-FTEII-T1033 cells stored in RNAlater solution (Ambion, Grand Island, NY, USA), previously grown in BMG medium supplemented with 0.1% (w/v) CaCl2, for 10 h at 30 °C and 250 rpm. The isolated RNA was treated with RQ1 RNase-free DNase (Promega, Madison, WI, USA). First-strand cDNA was synthesized by reverse transcription of the RNA using oligo(dT)15 primer and M-MLV reverse transcriptase, according to the manufacturer’s recommendations. PCR amplifications were performed using a PCR Multigene Mini Thermal Cycler (Labnet International Inc., Edison, NJ, USA) in a 25 µL reaction volume containing 0.5 µM of FTE1 and FTE2 primers, 0.2 mM dNTP’s each, 1X GoTaq reaction buffer, 1.0 U GoTaq DNA polymerase, and 2 µL of primary cDNA. Detection of β-actin transcripts was used as a positive RT-PCR control with 5ACT and 3ACT primers in the PCR step. A 30-cycle amplification program was used: 95 °C for 1 min, 50 or 56 °C (FTEII or β-actin transcript detection, respectively) for 1 min, and 72 °C for 1 min, with a first denaturation step at 95 °C for 1 min, and a final extension step at 72 °C for 5 min. The amplified products were visualized by agarose gel electrophoresis.

Proteins from the cell-free culture medium at 24 h culture of each strain were analyzed by SDS-PAGE to assess the migration shift of the Endo H-treated proteins in a Coomassie blue-stained 12% SDS-polyacrylamide gel. The cell-free culture medium was previously concentrated 100-fold and desalted by ultrafiltration at 4 °C using 10-kDa Amicon Ultra-4 filters (Millipore, MA, USA). Concentrated samples were then incubated with Endo H for 1 h at 37 °C in accordance with the manufacturer’s instructions.

Volumetric extracellular phytase activity was determined by measuring the phosphate released from sodium phytate as described previously (Guerrero-Olazarán et al. 2010) for the cell-free medium of 24 h culture, previously desalted using a PD-10 column (GE Healthcare Bio-Sciences Corp, Piscataway, NJ, USA) as described previously (Viader-Salvadó et al. 2010). One unit of phytase activity was defined as the amount of enzyme required to liberate 1 µmol of phosphate per min from sodium phytate under the assay conditions (pH 7.5, 37 °C).

Growth kinetics and activities of the regulatory elements (P1033-TAOX1 and P1033-T1033)

The two single-copy strains (KM71/P1033-FTEII-TAOX1 and KM71/P1033-FTEII-T1033) were grown in BMGly and BMGlc media supplemented with 0.1% (w/v) CaCl2 (initial OD600 of 1.0) for 24 h at 30 °C and 250 rpm. Samples were taken every 3 h up to 6 h of culture and later every 6 h to measure growth by the turbidimetry assay at OD600. Specific growth rates (µ) were determined by the slope of the natural log-linear regression of biomass concentration vs. time of the exponential phase (3 to 12 h) and statistically compared using a Student t-test with a significance level of 0.05.

The P1033 activities for the heterologous FTEII and endogenous 1033 genes in each single-copy strain were determined by reverse transcription-quantitative polymerase chain reaction (RT-qPCR) using the thermocycler Mx3005P QPCR system (Agilent Technologies, Santa Clara, CA, USA) and YPT1 as a normalizer gene. The GAPDH-transcript levels were also determined as transcript-level references. Cell samples were harvested by centrifugation (3000 g, 10 min, 4 °C) at 6, 12, and 18 h of culture and stored in RNAlater solution (Ambion, Grand Island, NY, USA) until used. Total RNA was obtained from the cell samples, the cDNA was synthesized and transcripts were amplified as described previously (Caballero-Pérez et al. 2021). PrimeTime qPCR Probe assays (Integrated DNA Technologies, Inc.) were used for the FTEII and GAPDH analyses (Herrera-Estala et al. 2022), and the SYBR Green DNA intercalator agent with 5q1033 and 3q1033 or 5qYPT1 and 3qYPT1 primers for the endogenous 1033 and YPT1 gene expression level determinations.

Extracellular product/biomass yield (Yp/x) was estimated for each culture at each sampling time, as the ratio of volumetric extracellular phytase activity (U per liter) to biomass concentration (g DCW per liter), which were estimated as above. The cell-free culture medium, obtained by centrifugation (3000 g, 10 min, 4 °C), was previously concentrated 5- to 50-fold and diafiltrated by ultrafiltration using 10-kDa Amicon Ultra-4 filters (Millipore, MA, USA) and 100 mM Tris-HCl (pH 8.5) buffer with 50 mM NaCl, 2% glycerol, and 5 mM CaCl2.

The transcript levels and Yp/x values were statistically compared using a Student’s t-test for independent samples and paired samples with a significance level of 0.05.

Results

P1033 and T1033 sequences

The P1033 sequence that was considered contained 176 nucleotides located upstream of the 1033 CDS. The TSS was identified at − 29 nucleotides upstream of the start codon (Fig. S1), the TATA-like box (TAAAAAGA) and the mammalian-type INR element (CTAATCG) were located at − 10 to − 17 and + 2 to + 8 nucleotides from the TSS, respectively (Fig. S2). The bioinformatics pipeline analysis of the P1033 sequence showed six TFBS and potential TF (i.e., Ecm22, Cat8, Rox1, Asg1, Yjl103c, and Stp2; Fig. S2).

The transcriptomic analysis showed that RNA reads aligned up to 107 nucleotides downstream of the 1033 CDS (Fig. S1), ending with a CAA that is a characteristic RNA cleavage and polyadenylation site for yeasts (Graber et al. 2002). The RNA cleavage and polyadenylation site was flanked by two A + T-rich sequences (ATATGTAT and ATTTATT) that are over-represented in downstream CDS sequences for yeasts (van Helden et al. 2000). Therefore, the considered T1033 sequence contained 157 nucleotides located downstream of the 1033 CDS.

Construction and selection of the K. phaffii KM71/P1033-FTEII-TAOX1 and KM71/P1033-FTEII-T1033 strains

The expression vectors pP1033-FTEII-TAOX1 and pP1033-FTEII-T1033 were constructed harboring a DNA coding sequence for a beta-propeller phytase FTEII (Viader-Salvadó et al. 2010) as the reporter gene, in-frame with the S. cerevisiae alpha-factor prepro-secretion signal and between P1033 and TAOX1 or P1033 and T1033, respectively (Fig. 1). PCR analysis of plasmid DNAs (pP1033-FTEII-TAOX1 and pP1033-FTEII-T1033), and the genomic DNAs isolated from KM71/P1033-FTEII-TAOX1 and KM71/P1033-FTEII-T1033 strains (His+ transformants), showed the expected bands of 1,357 and 870 bp, which confirmed the expression plasmid construction and the integration of the expression cassette into the K. phaffii genome.

Fig. 1
figure 1

pP1033-FTEII-TAOX1 and pP1033-FTEII-T1033 expression vectors. P1033, promoter of the 1033 gene; SS, alpha-factor prepro-secretion signal coding sequence; FTEII, gene coding for the beta-propeller phytase FTEII; TAOX1, AOX1 transcriptional terminator; T1033, transcriptional terminator of the 1033 gene; HIS4, K. phaffii wild-type gene coding for histidinol dehydrogenase; 3′AOX1, AOX1 downstream region; ColE1, Escherichia coli origin of replication; AmpR, ampicillin resistance gene; AatII, XhoI, NotI and Bsu36I restriction sites for cloning P1033 and T1033 sequences; SalI, restriction site for vector linearization before K. phaffii transformation by electroporation

Full size image

Cell-free culture media from 24 h cultures of 32 and 20 recombinant clones of each strain (KM71/P1033-FTEII-TAOX1 and KM71/P1033-FTEII-T1033, respectively), showed biomass concentrations ranging from 7.33 to 8.83 and 8.06 to 8.99 g DCW/L, extracellular protein concentrations from 5.93 to 17.20 and 5.31 to 8.52 mg/L, and extracellular protein/biomass yields from 0.7 to 2.0 and 0.6 to 1.0 mg/g, respectively (Fig. S3). The selected strain from each construction rendered the lowest protein production yield. The qPCR analysis of DNA from the selected strain of each construction confirmed the presence of one copy of the heterologous gene in the yeast genome.

P1033 and T1033 functionality

The RT-PCR assays for KM71/P1033-FTEII-TAOX1 and KM71/P1033-FTEII-T1033 cells in BMG medium with CaCl2 showed the expected 536-bp band of an FTEII-transcript fragment (Fig. 2a and b).

Fig. 2
figure 2

Agarose gels for RT-PCR assays from RNA of KM71/P1033-FTEII-TAOX1 cells (a) and KM71/P1033-FTEII-T1033 cells (b). Lane M1, molecular size marker, 100 bp DNA ladder (Promega). Lane M2, molecular size marker, 1 kb DNA ladder (Promega). Lanes 1 to 4, assays with the FTE1 and FTE2 primers. Lanes 1, negative control of reverse transcriptase step (without reverse transcriptase); lanes 2, PCR negative control (without cDNA); lanes 3, PCR positive control (pP1033-FTEII-TAOX1 plasmid); lanes 4, RT-PCR amplified product. Lanes 5 to 7, assays with the 5ACT and 3ACT primers. Lanes 5, negative control of reverse transcriptase step; lanes 6, PCR negative control; lanes 7, RT-PCR amplified product. c SDS-polyacrylamide gel of cell-free culture medium from BMG cultures, previously concentrated by ultrafiltration. Lane M, molecular mass marker. Lanes 1 and 2, proteins treated with and without Endo H glycosidase from the KM71/P1033-FTEII-TAOX1 culture, respectively. Lane 3, Endo H glycosidase. Lanes 4 and 5, proteins treated with and without Endo H glycosidase from the KM71/P1033-FTEII-T1033 culture, respectively. The arrow indicates N-deglycosylated recombinant FTEII

Full size image

The SDS-PAGE analysis for the cell-free culture medium of the KM71/P1033-FTEII-TAOX1 and KM71/P1033-FTEII-T1033 cultures showed the characteristic smear for phytase FTEII ranging from 39 to over 66 kDa (Fig. 2c, lanes 2 and 5) that shifted to a defined band of 39 kDa (Fig. 2c, lanes 1 and 4) after N-deglycosylation by Endo H which corresponds to the theoretical molecular mass for the phytase FTEII based on its amino acid sequence (Viader-Salvadó et al. 2010). This result confirms that the phytase FTEII produced in the K. phaffii expression system is secreted as a highly N-glycosylated protein (Viader-Salvadó et al. 2010). The cell-free media from the KM71/P1033-FTEII-TAOX1 and KM71/P1033-FTEII-T1033 cultures also showed phytase activities of 94.0 and 44.0 U/L, respectively.

These results (FTEII-transcript detection, phytase FTEII detection by SDS-PAGE, and enzyme activity) confirmed the correct functionality of the regulatory sequences and their combinations (i.e., P1033-TAOX1 and P1033-T1033).

Growth kinetics

Figure 3a shows the growth kinetics for the two strains using glucose or glycerol as the carbon sources. Cell growth increased exponentially from 3 to 12 h of culture for the four kinetic assays with µ values ranging from 0.203 ± 0.029 to 0.257 ± 0.001 h−1. No significant differences were seen in terms of µ between the same strain grown in glucose and glycerol or between different strains grown in the same carbon source. Although the four kinetic assays reached plateaus at 24 h of culture, the KM71/P1033-FTEII-TAOX1 and KM71/P1033-FTEII-T1033 cultures in glucose yielded 1.29 and 1.35 times more biomass concentrations, respectively, compared to the cultures grown in glycerol.

Fig. 3
figure 3

Growth kinetics (a) in BMGlc (●, ▲) and BMGly (■, ♦) media for the KM71/P1033-FTEII-TAOX1 (●, ■) and KM71/P1033-FTEII-T1033 (▲, ♦) strains. Points represent the mean from three independent kinetics experiments (coefficient of variation less than 5%). Relative transcript levels of heterologous FTEII (b), 1033 (c), and GAPDH (d) genes for the KM71/P1033-FTEII-TAOX1 (□, ■) and KM71/P1033-FTEII-T1033 ( , ) strains in BMGlc (□, ) and BMGly (■, ) media at different culture times. Data are presented as the mean ± standard error from three independent cultures. *Significant differences (p < 0.05) between glycerol- and glucose-grown cells from the same strain. #Significant differences (p < 0.05) between the two strains grown in the same carbon source

Full size image

Activities of the regulatory elements (P1033-TAOX1 and P1033-T1033)

The transcript levels of the FTEII gene (i.e., the reporter gene), which is regulated by P1033, were 1.4-times higher for glucose-grown KM71/P1033-FTEII-TAOX1 cells at the exponential growth phase, compared to the KM71/P1033-FTEII-T1033 cells (Fig. 3b). Nevertheless, when the two strains were grown in glycerol, no significant differences were seen between the FTEII-transcript levels at all sampling times. Moreover, the FTEII transcript-level profile for the KM71/P1033-FTEII-TAOX1 strain was linked to cell growth in glucose. Although the KM71/P1033-FTEII-T1033 strain showed an FTEII transcript-level profile linked to cell growth in glucose, it was not statistically significant. The FTEII-transcript levels were 1.3-times higher for glucose-grown, compared to glycerol-grown KM71/P1033-FTEII-TAOX1 cells at the exponential phase, and no significant differences were seen for the KM71/P1033-FTEII-T1033 cells in terms of the FTEII-transcript levels for cells grown in either of the carbon sources. The FTEII gene was more downregulated at the stationary phase for cells grown in glucose, compared to those grown in glycerol, since FTEII-transcript levels for KM71/P1033-FTEII-TAOX1 and KM71/P1033-FTEII-T1033 cells were 1.8-times and 3.6-times greater, respectively, at the stationary phase when cells were grown in glycerol, compared to glucose.

Transcript levels of the endogenous 1033 gene did not show significant differences between the two strains grown in glucose or glycerol (Fig. 3c). The 1033-transcript level profile was linked to cell growth in glycerol since the transcript levels increased significantly during the exponential growth phase (1.7-times and 1.4-times higher from 6 to 12 h of culture for KM71/P1033-FTEII-TAOX1 and KM71P1033-FTEII-T1033 cells, respectively). The same trend was observed in the 1033-transcript levels for glucose-grown cells and for glycerol-grown cells between 6 and 12 h of culture, though it was not statistically significant. The 1033-transcript levels were up to 1.4-times higher in glycerol-grown, compared to glucose-grown cells at 12 h of culture for both strains, though the increase was not statistically significant for the KM71/P1033-FTEII-TAOX1 strain. The 1033 gene was downregulated in both strains when the carbon source was depleted at the stationary phase.

Transcript levels of the endogenous GAPDH gene did not show significant differences between the two strains grown with glycerol as the carbon source, but in glucose-grown cells, the GAPDH-transcript levels were up to 1.8-times higher at the exponential growth phase for the KM71/P1033-FTEII-TAOX1 cells in comparison to the KM71/P1033-FTEII-T1033 cells (Fig. 3d). The GAPDH-transcript level profile remained linked to cell growth for the two strains and carbon sources, with a transcript-level increase during the exponential growth phase and a decrease when the cells reached the stationary phase. The GAPDH-transcript levels of the KM71/P1033-FTEII-TAOX1 strain at the exponential growth phase were 1.7-times higher for glucose-grown, compared to glycerol-grown cells, while GAPDH-transcript levels did not show significant differences for KM71/P1033-FTEII-T1033 cells grown with either of the two carbon sources.

FTEII-transcript levels were on average 2.8- and 2.1-times higher, and 2.0- and 2.2-times higher than the 1033-transcript levels for glucose-grown KM71/P1033-FTEII-TAOX1 and KM71/P1033-FTEII-T1033 cells, and for glycerol-grown KM71/P1033-FTEII-TAOX1 and KM71/P1033-FTEII-T1033 cells, respectively (Fig. 4a). During the stationary phase, the increased transcript-level ratio of FTEII to 1033 was enhanced in both carbon sources.

Fig. 4
figure 4

FTEII to 1033 transcript-level ratio (a), and GAPDH to FTEII (b) or 1033 (c) transcript-level ratio for the KM71/P1033-FTEII-TAOX1 ( □, ■) and KM71/P1033-FTEII-T1033 ( , ) strains in BMGlc (□, ) and BMGly (■, ) media at different culture times. Data are presented as the mean from three independent cultures. (D) Extracellular product/biomass yields (Yp/x) for the KM71/P1033-FTEII-TAOX1 (□, ■) and KM71/P1033-FTEII-T1033 ( , ) strains in BMGlc (□, ) and BMGly (■, ) media at different culture times. Data are presented as the mean ± standard error from three independent cultures. *Significant differences (p < 0.05) between glycerol- and glucose-grown cells from the same strain. #Significant differences (p < 0.05) between the two strains grown in the same carbon source

Full size image

GAPDH-transcript levels were on average 50- and 60-times higher than FTEII-transcript levels in the glucose cultures, but in the glycerol cultures they were 34- and 38-times higher in KM71/P1033-FTEII-TAOX1 and KM71/P1033-FTEII-T1033, respectively (Fig. 4b). The GAPDH-transcript levels were on average 142- and 125-times higher than the 1033-transcript levels for glucose-grown KM71/P1033-FTEII-TAOX1 and KM71/P1033-FTEII-T1033 cells, respectively (Fig. 4c). For glycerol-grown KM71/P1033-FTEII-TAOX1 and KM71/P1033-FTEII-T1033 cells, the transcript-level ratios of GAPDH to 1033 were 106 and 93, on average, respectively, at 6 h of culture.

Yp/x were, on average, 1.4-times higher for KM71/P1033-FTEII-TAOX1 cultures, compared to KM71/P1033-FTEII-T1033 cultures in glucose (Fig. 4d). This finding agrees with the higher (1.4-fold) FTEII-transcript levels of the KM71/P1033-FTEII-TAOX1 strain, compared to the KM71/P1033-FTEII-T1033 strain, both grown in glucose. Nevertheless, differences in the Yp/x were less pronounced among the glycerol-grown strains (1.1-times), which could be related to the absence of significant differences in the FTEII-transcript levels among the two strains grown in glycerol. Although the glucose-grown KM71/P1033-FTEII-TAOX1 cells showed higher FTEII-transcript levels compared to glycerol-grown cells, the Yp/x was 1.3-fold higher in glycerol than in the glucose cultures. Although no significant differences were seen in the FTEII-transcript levels of the KM71/P1033-FTEII-T1033 strain grown in glucose and glycerol, the Yp/x yield was 1.7-times higher in glycerol than in glucose cultures. These findings indicate that the translation and protein secretion pathway had a high impact on Yp/x when glycerol was used as a carbon source. The highest Yp/x in both strains occurred at the end of the exponential growth phase and the beginning of the stationary growth phase.

Discussion

Although a strong inducible or constitutive promoter is the first choice for expressing heterologous genes in K. phaffii, interest in weak promoters has recently increased due to their potential to produce host-toxic proteins (Ruth et al. 2010) or to simultaneously co-express several genes and maintain a metabolic balance in the host (Vogl et al. 2016; Jin et al. 2019; Duo et al., 2021). Few studies have been conducted on transcriptional terminators that can be used to tune gene expression, in comparison to the number of promoter studies (Ito et al. 2020). Therefore, in this work, we identified and functionally characterized for the first time the promoter (P1033) and transcriptional terminator (T1033) of a constitutive gene with weak, non-methanol-dependent transcriptional activity (i.e., the 1033 gene). This is the first report on the functional characterization of the promoter and terminal terminator of the K. phaffii 1033 gene that expands the available toolbox for tuning gene expression in K. phaffii.

The 5′UTR and 3′UTR of the 1033 gene are not annotated in the K. phaffii genomes available in the databases. For this reason, we analyzed the RNA-seq data to locate subregions upstream and downstream of the 1033 CDS where the RNA reads were not aligned. In the upstream region, the region where RNA reads were not mapped had a length of 136 nucleotides on average, while the inter-CDS region between the 1033 gene and the adjacent upstream gene is annotated with 176 nucleotides. Since these two regions are similar in size, we decided to use the sequence of the inter-CDS region as the P1033 sequence to be evaluated.

The P1033 contains a TATA-like sequence with two nucleotide changes from the TATA-box consensus sequence (TATAWAW), which is characteristic of constitutive promoters (Donczew and Hahn 2018). Moreover, the P1033 sequence harbors a mammalian-type INR element downstream of the TATA-like box. Mammalian-type INR sequences are present at, or adjacent to, the TSS of ∼ 40% of yeast promoters (Yang et al. 2007). The putative TSS was located at a CA sequence on the coding strand, i.e., pyrimidine-purine (PyPu) dinucleotide at the − 1 and + 1 sites of the TSS, with two adenines at a region of 7 nucleotides immediately upstream of the CA sequence. This TSS nucleotide sequence has been described for yeast species that use the scanning model for transcription initiation (Lu and Lin 2021). The P1033 sequence also contains a homopolymeric stretch of 21 deoxythymidine nucleotides (Ts) in the coding strand. Poly (dA:dT) sequences are overabundant in eukaryotic genomes (Segal and Widom 2009) and they are considered as an upstream promoter element in constitutive yeast promoters that act bidirectionally to activate the transcription of two adjacent unrelated genes transcribed in opposite directions (Struhl 1985, 1986), as is the case for the 1033 gene and its adjacent upstream gene (i.e., NOB1 gene).

Only 6 TFBS were identified in the P1033 sequence analysis, which is less than the 11 and 21 TFBS that we identified in the PAOX1 and PGAP sequences, respectively, from the same in silico analysis (unpublished data). Among the putative 6 TF for P1033, only one is shared between P1033 and PGAP (i.e., Cat8-1). Cat8 is a zinc cluster transcriptional activator necessary for the derepression of a variety of genes under non-fermentative growth conditions. In S. cerevisiae, Cat8 is one of the most important TF that activates genes of the gluconeogenesis pathway (Schüller 2003; Turcotte et al. 2010). K. phaffii has two putative Cat8 homologs (i.e., Cat8-1 and Cat8-2) that are involved in activating the ethanol assimilation pathway (Barbay et al. 2021). Moreover, Cat8-1 is necessary for activating genes in the glyoxylate cycle that contribute to the gluconeogenesis pathway. The identification of this TFBS in the P1033 sequence could be related to the putative function of the encoded protein from the 1033 gene (i.e., D-amino-acid oxidase) that catalyzes the oxidation of neutral D-amino acids to α-keto acids, which can enter the gluconeogenesis, glycolysis, and Krebs cycle pathways.

The transcriptome analysis downstream of the 1033 CDS was used to define the 3′URT end of the 1033 gene (i.e., 107 nucleotides downstream from the 1033 CDS end). Therefore, we considered the sequence of the 3′UTR region with 50 additional nucleotides as the sequence of T1033 to be evaluated, following suggestions in the literature for yeast transcriptional terminators (Curran et al. 2013).

The low strength of P1033 was initially observed with the transformant cultures from the two constructed strains since the range in protein/biomass yield was 5- to 23-times lower than those previously obtained for similar cultures with PGAP-TAOX1 clones using the same reporter gene (Herrera-Estala et al. 2022).

In previous works (Viader-Salvadó et al. 2010; Herrera-Estala et al. 2022) we saw that the expression cassette integration at the HIS4 locus by electroporation could lead to multiple insertion events at this locus. A K. phaffii strain that contains multiple integrated copies of the expression cassette usually yields higher heterologous gene-transcript and protein levels than a single-copy strain (Cereghino and Cregg 2000; Looser et al. 2015; Mombeni et al. 2020). For this reason, for a better comparison of the effect of the two genetic rearrangements of the expression cassette on the gene expression levels, we evaluated the gene dosage of several transformants by qPCR and selected single-copy clones of the two constructed strains (KM71/P1033-FTEII-TAOX1 and KM71/P1033-FTEII-T1033) to avoid the possible effect of gene dosage on the heterologous gene-transcript and protein levels and to compare it with the transcript levels of the endogenous 1033 and GAPDH genes in the two genetic rearrangements of the expression cassettes.

Generally, an intracellular reporter protein such as the green fluorescent protein is used to assess the strength of a promoter (Hartner et al. 2008; Qin et al. 2011; Prielhofer et al. 2013). Nevertheless, the fluorescent signal generated from this approach does not evaluate exclusively the promoter activity on the heterologous-gene transcription, otherwise, the transcription is measured together with the translation process. One of the greatest advantages of the K. phaffii expression system is its efficient protein secretion system. Hence, this host is often preferred for the extracellular production of recombinant proteins, which facilitates the downstream process. Consequently, we decided to evaluate the heterologous gene-transcript levels by RT-qPCR using an expression cassette harboring a reporter gene that codes for a protein secreted into the culture medium, and two combinations of promoter and transcriptional terminator pairs as regulatory DNA elements. With this approach, besides the promoter activity on the heterologous-gene transcription and the transcriptional terminator effect on transcript stability, we evaluated the impact of the regulatory elements on translation and protein secretion by measuring the extracellular product/biomass yield (Yp/x).

The differences in the P1033 transcriptional activity on FTEII between the two strains grown in glucose were not seen for the 1033 gene. In contrast, FTEII- and 1033-transcript levels did not show significant differences between the two strains in glycerol-grown cells. These results indicate that the TAOX1 and T1033 terminators did not contribute to differentiating the FTEII-transcript levels in glycerol-grown cells, but in glucose, TAOX1 increases the FTEII-transcript levels with respect to T1033, which shows the influence of the transcriptional terminator in the expression cassette on the heterologous gene-transcript levels. These findings are similar to those described previously (Ramakrishnan et al. 2020; Ito et al. 2020; Herrera-Estala et al. 2022), where the TAOX1 activity was greater than other endogenous transcriptional terminators, which was said to be due to the higher stability of the 3′UTR region and a higher mRNA half-life of the TAOX1. Our findings suggest that mRNA stability could also be dependent on the carbon source used for cell growth.

The endogenous 1033 gene was expressed with low transcript levels compared to the GAPDH gene, and its expression was modulated by cell growth and the carbon source. Both strains showed 1033-transcript levels that were 1.4-times higher in glycerol- than in glucose-grown cells during the exponential growth phase. Similar P1033 transcriptional activity on the 1033 gene was seen for the two strains in the same carbon sources. In the stationary phase, the 1033-transcript levels decreased, likely due to the absence of the carbon source. The decrease was more pronounced in glucose than in glycerol cultures. Although the FTEII and 1033 genes were driven by P1033 in both strains, the FTEII-transcriptional levels were higher than those of the 1033 gene, being FTEII to 1033-transcript level ratio highest in the KM71/P1033-FTEII-TAOX1 strain compared to the KM71/P1033-FTEII-T1033 strain in both carbon sources (Fig. 4). These findings indicate a competitive effect on FTEII and 1033 transcription when the same promoter was used two times in the same strain leading to a decrease on the transcript level of one of the genes, as was reported previously for PGAP-driven heterologous genes and endogenous GAPDH gene (Dou et al. 2021). This phenomenon was also enhanced by the higher stability of the FTEII transcripts harboring the 3′UTR region from the AOX1 gene than the 1033 transcripts harboring the 3′UTR region from the 1033 gene, as described previously for other 3′UTR regions (Ito et al. 2020).

Transcript levels of the endogenous GAPDH gene in the strain KM71/P1033-FTEII-TAOX1 were different according to growth phase and carbon source, with higher transcript levels in glucose-grown cells than in glycerol-grown cells during the exponential phase. This is consistent with PGAP, which has been traditionally cataloged as a strong constitutive promoter with greater activity at high specific growth rates (Looser et al. 2015) and in glucose cultures compared to glycerol cultures (Waterham et al. 1997). The GAPDH-transcript levels were higher in the KM71/P1033-FTEII-TAOX1 strain than in the KM71/P1033-FTEII-T1033 strain when cultures were grown in glucose. These results indicate that the presence of TAOX1 in the expression cassette renders a higher transcriptional activity for the endogenous GAP-promoter/GAP-terminator (PGAP-TGAP) combination in glucose cultures, in comparison to the T1033 in the expression cassette. The activity of PGAP trans-regulatory elements in the glucose cultures was likely affected by the promoter-terminator combination of the expression cassette. Therefore, the heterologous gene regulation driven by the P1033-TAOX1 combination in the expression cassette presented a lower competition with the endogenous GAPDH regulation (PGAP-TGAP) in glucose cultures, compared to the heterologous gene regulation driven by the P1033-T1033 combination. In contrast, no differences in GAPDH-transcript levels in glycerol-grown cells were seen between the two strains, which indicates that the regulation driven by the two heterologous promoter–terminator combinations competed similarly with the GAPDH regulation. Thus, we speculate that the regulatory DNA elements (i.e., promoter and transcriptional terminator) of the 1033 gene are regulated similarly to those of the GAPDH gene in glucose cultures, affecting the GAPDH regulation when two copies of the promoter-terminator pair (i.e., endogenous and heterologous) of the 1033 gene are present in the yeast genome, as was reported for PGAP-driven genes (Dou et al. 2021).

Although the highest heterologous gene transcription levels were attained in glucose-grown KM71/P1033-FTEII-TAOX1 cells, the highest Yp/x values were obtained in the glycerol cultures for the two strains. These results agree with previous reports that concluded that transcript levels by themselves are not sufficient to predict protein expression levels (Liu et al. 2016; Dou et al. 2021). Our results indicate that the promoter-terminator pair along with the carbon source affect the translation and/or protein secretion pathway. We found similar results with a PGAP-TAOX1 strain using the same reporter gene (Herrera-Estala et al. 2022), where the culture conditions that yielded low heterologous gene-transcript levels also yielded a high Yp/x, which was correlated with an upregulation of the KAR2 and PSA1-1 genes of the secretion pathway rather than with increased heterologous gene transcription. Furthermore, a culture strategy of a glycerol batch phase followed by a glucose-fed batch phase for a GAP-promoter K. phaffii system only increased the Yp/x by 1.1-times, compared to the glycerol-glycerol strategy (Garcia-Ortega et al. 2013), even though PGAP shows greater transcriptional activity in glucose than glycerol cultures (Waterham et al. 1997). Taken together, this data points to low heterologous gene-transcript levels along with glycerol cultures increase translation and/or protein secretion, at least for the reporter protein used in this work.

In conclusion, the findings indicate that the P1033 has 2–3% transcriptional activity of the PGAP activity and is tunable by cell growth and the carbon source. The combinations of the regulatory DNA elements from the K. phaffii 1033 and AOX1 genes (i.e., P1033-TAOX1 and P1033-T1033 pairs) contributed to a differentiated transcriptional activity of the heterologous and endogenous genes that were dependent on carbon source, which shows the influence of the transcriptional terminator in the expression cassette on the heterologous and endogenous gene-transcript levels. The promoter-terminator pair and the carbon source affected the translation and/or protein secretion pathway. Moreover, low heterologous gene-transcript levels along with glycerol cultures increased translation and/or protein secretion. The P1033 in combination with the TAOX1 or T1033 could be used to produce host-toxic proteins, co-express genes of the secretory pathway, and maintain a metabolic balance in metabolic engineering or synthetic biology applications.