2 Institute of Plant Physiology and Ecology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai 200032, China
Rapamycin is a macrocyclic polyketide antibiotic exhi-biting immunosuppressive, anticancer, and antifungal activities [1-4]. It has a typical polyketide structure, in which DHCHC (4, 5-dihydroxycyclohex-1-ene carboxylic acid) is loaded as the starting unit at the beginning of its synthesis. Seven malonyl-CoA molecules and seven methylmalonyl-CoA molecules are in turn condensed to form a polyketide chain catalyzed by polyketide synthases (PKS) RapA, RapB and RapC. Then, one molecule of L-pipecolic acid is incorporated and condensed with the polyketide chain by a non-ribosomal peptide synthase encoded by rapP gene, to form rapamycin macrolactone. After a series of modifications such as oxidations and methy-lations, rapamycin is generated [5-6].
A variety of rapamycin derivatives have been obtained by mutasynthesis, precursor directed biosynthesis, and chemical synthesis. For example, three new compounds, including prolylrapamy-cin , 20-thiarapamycin and 15-deoxo-19-sulfo-xylrapamycin  have been synthesized by replacing L-pi-pecolate with its structure analogues. Since the structure of macrolactone is determined by distinct catalytic domains within corresponding modular polyketide synthases (PKSs), the rapamycin analogs may be accessible through genetic engineering or reprograming the domains of PKS.
As shown in Fig. 1, FK506/FK520 and rapamycin are similar in chemical structure as well as their bioactivities. There is a double bond between C-27 and C-28 in FK506/FK520. However, a single bond exists at the corresponding position (between C-35 and C-36) of rapamycin. By analyzing the PKS modules (encoded by genes fkbB and rapA respectively) responsible for the first chain extension cycle of FK506 and rapamycin, an additional ER domain has been found to be present in the first module of rapA in rapamycin synthetic cluster, compared to that corresponding module of fkbB [6, 9]. In the present study, the conserved motif of active center in the ER domain of the first module of Ac_rapA was inactivated by site mutation with an intention to synthesize a new rapamycin analog with a double bond between C-35 and C-36. The resultant mutant was constructed and a novel rapamycin analog was obtained.
Bacterial strains, plasmids and primers used in the present study are listed in Tables 1 and 2. Actinoplanes sp. N902-109 was deposited at the National Institute of Bioscience and Human-Technology, Agency of Industrial Science and Tech-nology [Tokyo, Japan] with accession number FERM BP-3832.
E. coli was grown in LB medium (trptone 1%, yeast extract 0.5%, NaCl 1%) at 37 ℃. Actinoplanes sp. N902-109 was cultivated in Isp2 media (per liter: malt Extract 10 g; glucose 4 g; yeast extract 4 g; pH 7.0, adding agar 20 g for solid medium) at 28-30 ℃. M-Isp4 medium (per liter:soybean meal 5 g; mannitol 5 g; starch 5 g; tryptone 2 g; yeast extract 1 g; NaCl 1 g; (NH4)2SO4 2 g; K2HPO3 1 g; CaCO3 2 g; trace element solution 100 μL; pH 7.0; agar 20 g. Inorganic trace element solution: ZnSO4 0.001%; FeSO4 0.001%; and MnSO4 0.001%) was used for conjugant selection. The seed medium for rapamycin production in batch cultures contained the following (per liter): soluble starch 30 g; glucose 20 g; dry yeast 5 g; corn syrup powder 5 g; peanut meal 5 g; and CaCO3 2 g; pH 7.0. Fermentation medium contained the following (per liter): soluble starch 30 g; glucose 40 g; corn syrup powder 8 g; cotton seed protein powder 8 g; CaCO3 3 g; soak enemy 2 g; MgSO4·7H2O 0.002 5 g; CoCl2·6H2O 0.001 g; VB1 0.2 g; VB12 0.000 2 g; and folic acid 0.025 g; pH 6.9.Reagents and materials
Restriction endonucleases, ligase and T4 DNA polymerase were purchased from Thermo Scientific (Waltham, MA USA) and TaKaRa (Dalian, China), respectively. The high fidelity DNA polymerase KOD NEO plus was a product of Toyobo (Osaka, Japan); 2-mm-gap electroporation cuvette from Harvard Apparatus (Holliston, MA, USA) and Gene Pulser Xcell microbial electroporation system (Bio-Rad, USA) are used for electro-poration. Agilent 1200 series HPLC (Santa Clara, CA, USA) equipped with NaVa-PaK C8 column (3.9 mm × 150 mm, 5 μm, Waters, Milford, MA, USA) is used for products isolation. Q-TOF Micro-mass mass spectrometry (Waters, Milford, MA, USA) and NMR Inova-400 (Varian, Palo Alto, CA, USA) are used for chemical structure determination.Prediction of polyketide synthase functional domains
The complete sequence of rapamycin biosynthetic gene cluster has been reported previously (deposited under the GenBank accession number GQ166664) [13-14]. The prediction of functional domains in polyketide synthase was based on online tools: http://nrps.igs.umaryland.edu/nrps/.Plasmids construction
Genomic DNA extraction of Actinoplanes sp. N902-109 was performed according to the procedures reported previously .
The pLYERIA was constructed by the following procedure. Plasmid pLYZL102 was lineated by HindⅢ and XbaⅠ. The homolo-gous arms (HA) were amplified from Actinoplanes sp. N902-109 genomic DNA with primers ERUAs/ERUAas and ERDAas/ERDAs, respectively. The resultant upstream and downstream HAs were digested with NheⅠ/XbaⅠ and NheⅠ/ HindⅢ, and ligated with linear vector pLYZL102 to generate pLYERIA. The introduction of NheⅠ site facilitated identi-fication of the mutant from wild type RAC001 among double crossover conjugants (Fig. 2A).
Ⅰ-SceI endonuclease expressing plasmid pSETSI was constructed as shown in Fig. 2B. The pSET152 was digested with pShAI to remove apramycin resistant gene (encoded by aac(3)Ⅳ). The spectinomycin resistant gene (encoded by aadA) fragment was excised from pIJ788, blunted, and ligated with above linear pSET152 vector to get intermediate plasmid pSETspc. The promoter of apramycin Paac was amplified from pIJ773 with primers paacs and paacas followed by BamHI/ NdeⅠ digestion. Ⅰ-SceI gene was cut out with NdeⅠ and EcoRI from pLU101. pSETspc was lineated with BamHI and EcoRI, and then used as a vector. Paac and Ⅰ-SceI fragments were ligated with pSETspc vector to generate pSETSI.Chromosomal Inactivation of the enoylreductase in Actinop-lanes sp. N902-109
RAC001 was grown on Isp2 agar at 28-30 ℃ for 5 days. The mycelia were scraped and dispersed by vortex in sterile water with glass beads. The mycelia suspension was filtered by passing through a sterile syringe filled with cotton plug. After being centrifuged for 10 min at 7 000 r·min-1, mycelia were re-suspended with 1.0 mL of sterile water to give myce-lium suspension.
The pLYERIA was delivered into Actinoplanes sp. N902-109 (RAC001) by conjugation helped with donor E.coli ET12567/pUZ8002. Then the conjugation was carried out acco-rding to a classic method , except the mycelia suspen-sion of RAC001, replacing the spores, and used as a receptor in conjugation procedure. The conjugants were selected with 30 μg·mL-1 of apramycin on M-Isp4 agar medium instead of MS.
The single crossover event of the conjugant was identified by PCR using primers apr-V-R and ERvas. pSETSI was subsequently transferred into the single crossover mutant (named as ERSC) by conjugation. The pSETSI integration on the chromosome (named as ERSCI) was selected with 50 μg·mL-1 of spectinomycin for conjugants. ERSCI was grown in Isp2 liquid medium for subculture, and then streaked on Isp2 agar medium. The grown-up clones were dotted on Isp2 agar medium with or without 30 μg·mL-1 of apramycin. The double crossover event of the conjugant was identified by their sensitivity to apramycin. Those conjugants sensitive to apramycin were amplified with primers ERvs and ERvas. The PCR products were digested with NheⅠ to distinguish the enoylre-ductase mutants from the wild type ones. The nucleotides sequence of mutated site of enoylreductase domain was confirmed by PCR product sequencing.Fermentation of RAC001 and its mutants
After growing on seed slant medium at 28 ℃ for 7-9 days, RAC001 and its mutants were inoculated and cultured in flasks with 30 mL of seed medium for 4-5 days at 220 r·min-1. The seed cultures were transferred to second seed flasks with inoculum 10% (V/V) and grown for 3-4 days. Afterwards, the second seeds were transferred to fermentation flasks with inoculum 10% (V/V) and grown for 9 days.Extraction and purification of SIPI-rapxin
Fermentation broths were mixed with diatomite (1/40 volume) and centrifuged to recover the mycelia with diatomite that were extracted with equal volume acetone. The acetone extract was absorbed with resin, and eluted by ethyl acetate. The ethyl acetate extract was dehydrated with Na2SO4 and concentrated under vacuum at 40 ℃. The resultant slurry was loaded on silica gel column, which was eluted by ethyl acetate and the collection solutions containing SIPI-rapxin were combined and concentrated under vacuum at 35 ℃. Finally, the resultant compound sample was dissolved in methylene chloride and purified by preparative HPLC.HPLC, ESI-MS and nucleus magnetic resonance (NMR) measurements
The sample, extracted from crude fermentation culture with equal volume acetone, was centrifuged at 12 000 r·min-1 for 5 min. 20 μL of filtered supernatant was analyzed by HPLC (Agilent 1200 series HPLC (Santa Clara, CA, USA) equipped with NaVa-PaK C8 column (Waters, Milford, MA, USA) under the following conditions: mobile phase as acetonitrile: water (65 : 35), flow rate at1.0 mL·min-1, column temperature st 50 ℃, and UV detection wavelength at 278 nm. ESI-MS was performed on Q-TOF Micro-mass mass spectro-metry (Waters, Milford, MA, USA) and NMR spectrum analysis was carried out on Inova-400 (Varian, Palo Alto, CA, USA). 1H NMR and 13C NMR were performed at 400 MHz.Immunosuppressive activity assay
The immunosuppressive activities of compounds, including SIPI-rapxin, 27-O-demethylrapaycin, rapamycin and FK-506, were determined in MLR . Human fresh lymphocytes were separated from the heparinized blood of normal volunteers using density gradient centrifugation (Human cells were used following approval by the authors' institutional Human Ethics Review Committee). Stimulator cells (5 × 104) were prepared in the same way and stored in liquid nitrogen. Fresh lymphocytes were cultured in the presence of thawed stimulator cells (5 × 104) with or without compounds at 37 ℃ in the 5% CO2-humidified air, water-jacketed incubator for 5 days. Each well was treated with 3H-thymidine for 18 h. The cells were harvested onto glass filters, dried, and counted in liquid scintillation cocktail.Results Prediction of polyketide synthase functional domains
The predicted functional domains of Ac_rapA are shown in Table 3. It was composed of a loading and four elongation modules. The grey squares indicate the predicted inactive domains. According to the biosynthetic procedure of rapamycin/ FK506 and the corresponding domain arrangement of polyketide synthase, the ER domain in Module 1 (closed square) was predicted to be responsible for the reduction of carbon bond between C-35 and C-36.
ER domain editing plasmid, pLYERIA, which was consisted of the nucleotide mutations, corresponding with the amino acids mutations in active center of enoylreductase, and the homologous arms of ER domain, as well as the recognition sequence of Ⅰ-SceI on its backbone (Fig. 2A), weredelivered into RAC001. The resultant conjugants carrying the pLYERIA which had integrated into chromosome by homologous recombination could survive from apramycin selection (Fig. 3A).
A total of two single crossover conjugants (SC) were obtained and confirmed that pLYERIA inserted at ER domain through single crossover event (Fig. 3B). Ⅰ-SceI was subsequently over-expressed in single cross conjugant SC by integrating pSETSI into attB site in SC chromosome. A total of 7 apramycin-sensitive conjugants (A1-A7) were identified by PCR using primers ERvs and ERvas with wild type genome DNA (AC) as a control. The NheⅠ digestion of PCR product showed 6 out of 7 strains yielded two bands (1.9 and 2.9 kbp), demonstrating that the mutations were introduced into the position as expected (Fig. 3B). The sequencing of the mutation region confirmed the precise mutation in active center, i.e., the amino acid residue of Ala-Gly-Gly had been mutated to Ala-Ser-Pro (Fig. 3C).Isolation, purification, and identification of SIPI-rapxin
Enoylreductase mutant RAC027 and wild type RAC001 were cultivated and the broths were analyzed by HPLC. The results showed that the production of rapamycin decreased drastically and a new compound presented in the broth extract of RAC027, compared with RAC001 (Fig. 4A). Moreover, the unknown compound exhibited a similar UV spectrum to that of rapamycin, suggesting that a new rapamycin analog (named as SIPI-Rapxin subsequently) was synthesized by the mutant RAC027, not RAC001 (Fig. 4B). Based on the result of ESI-MS, the molecular weight of SIPI-rapxin was m/z 897 and m/z 16 less than that of rapamycin, which was probably due to the formation of a double bond (reducing two hydrogen protons 2H, m/z 2) and the loss of one methylene group (reducing CH2, m/z 14) compared to rapamycin (Fig. 4C). Therefore, its molecular formula of SIPI-Rapxin was C49H73NO13.
The 1H and 13C NMR spectra of SIPI-rapxin and 27-O-de-me-thylrapamycin were analyzed and compared. Each hydrogen and carbon in chemical structure was assigned as shown in Table 4. Most chemical shifts of hydrogen and carbon in SIPI-rapxin and 27-O-demethylrapamycin were similar. For example, the 1H NMR and 13C NMR spectra of SIPI-rapxin gave two 3H signals of methoxyl groups (δ 3.13 and 3.37) and two carbon signals of methoxyl groups (δ 55.8 and 56.4), which were similar with those of 27-O-Demethylrapamycin (δ 3.12 and 3.40 for 3H signals, δ 55.9 and 56.8 for carbon signals) at C-16, C-39. In addition, an extra double bond (δ 131.5 and 132.8), which was absent in 27-O-demethylrapamycin, occurred between C-35 and C-36 in SIPI-rapxin. Hence, SIPI-rapxin was a new analog of 27-O-demethylrapamycin. Its structure was determined as 35, 36-didehydro-27-O-de-methylrapamycin (Fig. 5).
The immunosuppressant activities of rapamycin and its analogs were analyzed using the mixed lymphocyte reaction (MLR) method and the results are shown in Table 5. The new compound SIPI-rapxin was more potent than rapamycin and 27-O-demethylrapamycin, and comparable to FK-506, implying that the double bond at C-35 and 36 was critical for MLR activity.
Combinatorial biosynthesis has been extensively applied in new structure compound discovery and structural analogue synthesis by manipulating the synthetic genes in the chromosome of producers, evading the bottleneck of chemical synthesis, such as stereoselectivity . Polyketide synthase is a kind of multinodular and multifunctional protein, which is composed of functional domains responsible for each step of product synthesis. Combinatorial biosynthesis strategy has been used to achieve erythromycin analog synthesis in Saccharopolyspora erythraea by engineering the polyketide synthase eryAII which functions in building erythromycin polyketide skeleton .
Although Actinoplanes sp. N902-109 is a second rapamycin producer after Streptomyces hygroscopicus ATCC29253, none of its genetic manipulation has been reported previously. In this work, assisted with the Ⅰ-SceI inducing DSB on chro-mosome, non-replicating plasmid was used to inactivate enoylreductase domain by homologous recombination, achie-ving the production of a novel rapamycin analog, 35, 36-didehydro-27-O-demethylrapamycin. The double bond presented in this compound demonstrated that the ER domain of Module 1 in Ac_rapA was responsible for the reduction of the double bond between C-35 and C-36 in rapamycin synthesis. Unexpectedly, a methoxyl group of rapamycin was also replaced by a hydroxyl group at C-27 of SIPI-rapxin. It has been reported that, in S. hygroscopicus, RapN is responsible for oxidation of C-27 and RapQ for subsequent methylation of this hydroxyl group in rapamycin synthesis . The functional analysis of the rapamycin synthetic cluster in Actinoplanes sp. N902-109 has uncovered two methyltra-nsferase homologs of Act_rapM, annotated as Act_rapM1 and Act_rapM2. However, no homolog of RapQ has been discovered thus far . We supposed that Act_rapM1 and Act_rapM2 were the counterparts of RapM and RapQ, which were in charge of the methylation of hydroxyl groups at C-27 and C-16. Subsequent PCR verification proved the presence of intact Act_rapM1 and Act_rapM2 in chromosome of all the ER mutants, excluding the possibility that demethylation of C-27 due to the absence of methyltransferases (data not shown).
It was notable that 27-O-demethylrapamycin usually exists in the broth of wild type Actinoplanes sp. N902-109, probably due to inefficient catalytic ability or poor substrate specificity of the methyltransferase. Because the elongation and cyclization of polyketide chain took place ahead of the post-mo-dification, including the methylation of C-27, there was a possibility that the double bond between C-35 and C-36 resulted in the variation of substrate spatial structure of methyltransferase, further decreasing the catalytic efficiency or specificity of the methyltransferase.
SIPI-rapxin and FK-506 with a similar double bond in structures showed approximately 2-fold more potency of immunosuppressant activities, compared to rapamycin, suggesting that the double bond between C-35 and C-36 was essential for immunosuppressant activity.
It has been proven that Rapamycin contains two distinct functional regions, which bind to FKBP (FK506-binding protein 12) and FRB (FKBP12-rapamycin binding) domains, respectively [21-22]. Several rapalogs which have the pipecolate ring substituted with its analogs in FKBP region lost activities significantly [8, 23], which suggest the pipecolate ring is crucial for the immunosuppressive activity. The FRB region is a potential target for further structure modification. The structural changes within this region, such as the elongation unit interchanges, will generate new compounds with different activity. It is also a huge challenge because the relevant structural changes will probably break the binding and recognition between the substrate and catalytic site. On the other hand, the loading molecule DHCHC is a promising target. Everolimus, Temsirolimus and Deforolimus which have been approved by FDA for clinical use are all rapamycin derivatives with different modifications on DHCHC module. However, the complete biosynthetic pathway of DHCHC remains unclear. Further investigation is needed and will contribute to the combinatorial biosynthesis of rapalogs, especially those with DHCHC modification.
In summary, we accomplished a modification on macrolactone of rapamycin through combinatorial synthesis strategy, which was inaccessible for chemical synthesis method, demonstrating its innovative and promising prospect for polyketide analog synthesis.
Martel RR, Klicius J, Galet S. Inhibition of immune-response by rapamycin, a new antifungal antibiotic[J]. Can J Physiol Pharmacol, 1977, 55(1): 48-51. DOI:10.1139/y77-007
Zeng Z, Sarbassov DD, Giles FJ, et al. Rapamycin analogs reduce mTORC2 signaling and inhibit AKT activation in AML[J]. Blood, 2006, 109(8): 3509-3512.
Metcalf C, Boyce B, Lu Y, et al. Bone-targeted rapamycin analogs inhibit osteolysis and tumor growth in a PC-3-GFP intratibial bone cancer model in nude mice[J]. Proc Am Assoc Cancer Res Ann Meet, 2006, 47: 263-266.
Metcalf C, Boyce B, Lu Y, et al. A tissue selective rapamycin analog inhibits MDA-MB-231 breast cancer cell-induced osteolytic bone metastases in nude mice[J]. Proc Am Assoc Cancer Res Ann Meet, 2006, 47: 1140-1141.
Park SR, Yoo YJ, Ban YH, et al. Biosynthesis of rapamycin and its regulation:past achievements and recent progress[J]. J Antibiot (Tokyo), 2010, 63(8): 434-441. DOI:10.1038/ja.2010.71
Schwecke Tea. The biosynthetic gene cluster for the polyketide immunosuppressant rapamycin[J]. Proc Natl Acad Sci USA, 1995, 92(17): 7839-7843. DOI:10.1073/pnas.92.17.7839
Vignot SF, Aguirre D, Raymond E. mTOR-targeted therapy of cancer with rapamycin derivatives[J]. Ann Oncol, 2005, 16(4): 525-537. DOI:10.1093/annonc/mdi113
Ritacco FV, Graziani EI, Summers MY, et al. Production of novel rapamycin analogs by precursor-directed biosynthesis[J]. Appl Environ Microbiol, 2005, 71(4): 1971-1976. DOI:10.1128/AEM.71.4.1971-1976.2005
Aparicio JF, Molnar I, Schwecke T, et al. Organization of the biosynthetic gene cluster for rapamycin in Streptomyces hygroscopicus:Analysis of the enzymatic domains in the modular polyketide synthase[J]. Gene, 1996, 169(1): 9-16. DOI:10.1016/0378-1119(95)00800-4
Nishida H, Sakakibara T, Yamauchi Y, et al. Rapamycin producer[M]. Pfizer Inc., 1997, 428.
Bierman M, Logan R, Obrien K, et al. Plasmid cloning vectors for the conjugal transfer of DNA from Escherichia coli to Streptomyces Spp[J]. Gene, 1992, 116(1): 43-49. DOI:10.1016/0378-1119(92)90627-2
Gust B, Challis GL, Fowler K, et al. PCR-targeted strepto-myces gene replacement identifies a protein domain needed for biosynthesis of the sesquiterpene soil odor geosmin[J]. Proc Natl Acad Sci USA, 2003, 100(4): 1541-1546. DOI:10.1073/pnas.0337542100
Huang H, Lu XJ, Hu HF. Analysis and determination of complete genome sequence of Actinoplanes sp. N902-109[J]. Chin J Antibiot, 2015, 40(3): 171-177.
Huang H, Ren SX, Yang S, et al. Comparative analysis of rapamycin biosynthesis clusters between Actinoplanes sp. N902-109 and Streptomyces hygroscopicus ATCC29253[J]. Chin J Nat Med, 2015, 13(2): 90-98.
Kieser T, Bibb MJ, Buttner MJ, et al. Practical Streptomyces Genetics:A Laboratory Manual[M]. Norwich, United Kingdom: John Innes Foundation, 2000.
Nishida H, Sakakibara T, Aoki F, et al. Generation of novel rapamycin structures by microbial manipulations[J]. J Antibiot (Tokyo), 1995, 48(7): 657-666. DOI:10.7164/antibiotics.48.657
Box SJ, Shelley PR, Tyler JW, et al. 27-O-Demethylrapamycin, an immunosuppressant compound produced by a new strain of streptomyces-hygroscopicus[J]. J Antibiot (Tokyo), 1995, 48(11): 1347-1349. DOI:10.7164/antibiotics.48.1347
Wong FT, Khosla C. Combinatorial biosynthesis of polyketides-a perspective[J]. Curr Opin Chem Biol, 2012, 16(1-2): 117-123. DOI:10.1016/j.cbpa.2012.01.018
Stefano D, Paul J, Marianna J, et al. An erythromycin analog produced by reprogramming of polyketide synthesis[J]. Proc Natl Acad Sci USA, 1993, 90(15): 7119-7123. DOI:10.1073/pnas.90.15.7119
Graziani EI. Recent advances in the chemistry, biosynthesis and pharmacology of rapamycin analogs[J]. Nat Prod Rep, 2009, 26(5): 602-609. DOI:10.1039/b804602f
Clardy J. The chemistry of signal transduction[J]. Proc Natl Acad Sci, 1995, 92(1): 56-61. DOI:10.1073/pnas.92.1.56
Kissinger CR, Parge HE, Knighton DR, et al. Crystal structures of human calcineurin and the human FKBP12–FK506–calci-neurin complex[J]. Nature, 1995, 378(6557): 641-644. DOI:10.1038/378641a0
Khaw LE, Bohm GA, Metcalfe S, et al. Mutational biosy-nthesis of novel rapamycins by a strain of Streptomyces hygroscopicus NRRL 5491 disrupted in rapL, encoding a putative lysine cyclodeaminase[J]. J Bacteriol, 1998, 180(4): 809-814.