A Novel FadL Homolog, AltL, Mediates Transport of Long-Chain Alkanes and Fatty Acids in Acinetobacter venetianus RAG-1

ABSTRACT

Due to the barrier effect of lipopolysaccharide (LPS) in the outer membrane of Gram-negative bacteria, transporters are required for hydrophobic alkane uptake. However, there are few reports on long-chain alkane transporters. In this study, a potential long-chain alkane transporter (AltL) was screened in Acinetobacter venetianus RAG-1 by comparative transcriptome analysis. Growth and degradation experiments showed that altL deletion led to the loss of n-octacosane utilization capacity of RAG-1. To identify the function of AltL, we measured the existence and accumulation of alkanes in cells through the constructed alkane detection system and isotope transport experiment, which proved its long-chain alkane transport function. Growth experiments using different chain-length n-alkanes and fatty acids as substrates showed that AltL was responsible for the transport of (very) long-chain n-alkanes (C₂₀ to C₃₈) and fatty acids (C_18A to C_28A) and was also involved in the uptake of medium-chain n-alkanes (C₁₆ to C₁₈). Subsequently, we analyzed the distribution of AltL in bacteria, and found that AltL homologs are widespread in Gamma-, Beta-, and Deltaproteobacteria. An AltL homolog in Pseudomonas aeruginosa was also identified to participate in long-chain alkane transport by a gene deletion and growth assay. We also found that overexpression of altL in Pseudomonas aeruginosa enhanced the degradation of C₁₆ to C₃₂ n-alkanes. In addition, structure analysis showed that AltL has longer extracellular loops than other FadL family members, which may be involved in the binding of alkanes. These results showed that AltL is a novel transporter and that it is mainly responsible for the transport of long-chain n-alkanes and (very) long-chain fatty acids and has broad application potential.

IMPORTANCE Petroleum pollution has caused great harm to the natural environment, and alkanes are the main components of petroleum. Many Gram-negative bacteria can use alkanes as carbon and energy sources, which is an important strategy for oil pollution remediation. Alkane uptake is the first step for its utilization. Hence, the characterization of transport proteins is of great significance for the recovery of oil pollution and other potential applications in industrial engineering bacteria. At present, some short- and medium-chain alkane transporters have been identified, but stronger hydrophobic long-chain alkane transporters have received little attention. In this study, the broad-spectrum transporter AltL, identified in RAG-1, makes up for the lack of research on the transport of long-chain alkanes and (very) long-chain fatty acids. Meanwhile, the structural features of longer extracellular loops might be related to its unique transport function on more hydrophobic and larger substrates, indicating it is a novel type alkane transporter.

INTRODUCTION

Alkanes, found in polluted and unpolluted sites, are ubiquitous in the environment. Every year, millions of tons of alkanes from petroleum are released into the biosphere through natural oil seeps and human activities (1, 2), and alkanes produced by plants, insects, and microbes can also diffuse into the environment (3). As a central part of the global alkane cycle, microorganisms have evolved various metabolic pathways to utilize these energy-rich substances (4–6), in which the uptake of alkanes is the initial and prerequisite step.

Alkanes are a rich energy source but also potentially toxic (7). Nonpolar alkane molecules are highly hydrophobic and have low water solubility, and these properties strengthen with increasing carbon chain length. The outer membrane (OM) of Gram-negative bacteria is surrounded by a lipopolysaccharide (LPS) layer (8, 9), which hinders the uptake of hydrophobic substrates, and dedicated membrane proteins are required to facilitate transport of alkanes across the OM. AlkL in Pseudomonas putida GPo1 functions as a passive importer of short- and medium-chain alkanes (C₁₂ to C₁₆) (10, 11). It belongs to the OmpW family, with a typical eight-stranded β-barrel structure and a narrow pore diameter (12). AupA is a transporter in Marinobacter hydrocarbonoclasticus SP17 involved in the binding and incorporation of micellar n-hexadecane (C₁₆) rather than accessible alkanes (13). At present, a few OM proteins are known to participate in long-chain alkane transport. In Alcanivorax dieselolei B5, three OM proteins named OmpT-3, OmpT-2, and OmpT-1 were identified to participate in the transport of C₈ to C₁₂, C₁₆ to C₂₄, and C₂₈ to C₃₆ n-alkanes, respectively (14). Other proteins responsible for long-chain alkane transport have not been reported.

FadL family members mediate the transport of hydrophobic substances across the OM (8, 15). FadL_Ec in Escherichia coli, the archetype of the family, is responsible for long-chain fatty acid transport (16, 17). TodX in P. putida F1 and TbuX in Ralstonia pickettii PKO1 mediate the transport of monoaromatic hydrocarbon (18, 19). Recently, FadL_Ec was also confirmed to participate in the import of alkanes (C₈ and C₁₀) into E. coli (20). These FadL family members share similar 14-stranded β-barrel structures with an interior occluded by an N-terminal hatch and an inward-pointing kink in β-strand S3 that creates a lateral opening in the transmembrane barrel, implying a lateral diffusion transport mechanism (9, 21, 22). The substrate specificity of FadL family members has been determined for FadL_Ec and TodX/TbuX (22).

Acinetobacter venetianus can utilize alkanes (23), and this species has been proposed as a model for studying alkane degradation mechanisms and a potential platform for industrial applications (24). RAG-1, a typical A. venetianus strain, exhibits superior emulsification and alkane degradation performance compared to other strains in the genus Acinetobacter (24, 25). Thus, RAG-1 was selected to explore the transport mechanism of long-chain alkanes.

In this study, the novel alkane transporter AltL in RAG-1 was discovered and found to transport medium- and long-chain alkanes (medium-chain, C₁₆ to C₁₈; long-chain, C₂₀ to C₃₈) and (very) long-chain fatty acids (long-chain, C_18A; very long-chain, C_20A to C_28A). Structural analysis revealed a typical β-barrel structure but longer extracellular loops than in other FadL family members, and L4 might be the decisive domain for alkane binding. Phylogenetic and functional analyses revealed the widespread distribution of AltL homologs in Gamma-, Beta-, and Deltaproteobacteria and potential functions in other bacteria, suggesting that AltL homologs might be ubiquitous alkane transporters in these species. These proteins may facilitate alkane cycling by allowing these bacteria to effectively obtain alkanes from the environment.

RESULTS

Deletion of altL disrupts the utilization of C₂₈.

To explore potential transporters involved in long-chain alkane uptake, comparative transcriptome analysis was conducted by culturing RAG-1 with n-octacosane (C28) as the sole carbon source, compared with sodium acetate (SA). Four genes (F959_RS02655, F959_RS06225, F959_RS06795, and F959_RS14525) encoding a predicted TonB-dependent receptor, outer membrane protein E, outer membrane protein OmpW, and hypothetical outer membrane protein, were upregulated 4.66-, 4.32-, 2.68-, and 2.66-fold, respectively (see Table S1 in the Supplemental Material). Quantitative real-time PCR (qRT-PCR) confirmed the RNA sequencing (RNA-seq) results (Table S1).

To investigate whether these genes are involved in C₂₈ metabolism, four single-gene deletion mutants were constructed, and growth curves showed that, compared with the wild-type (WT) strain, single-gene deletions had no significant effects on growth in lysogeny broth (LB) medium (Fig. S1). When cultured with C₂₈ as the sole carbon source, the three mutants (ΔF959_RS06795, ΔF959_RS02655, and ΔF959_RS06225) could grow and retained their alkane degradation ability, whereas deletion of F959_RS14525 basically completely abolished growth and degradation on C₂₈ (Fig. 1A and B). The gene (F959_RS14525) was named long-chain alkane transporter (AltL). To exclude the polar effect caused by gene deletion, ΔF959_RS14525 (ΔaltL) was complemented with altL using a recombinant plasmid, which restored the growth and degradation ability of ΔaltL on C₂₈, in contrast to the vector control altL/P (Fig. 1A and B). These results suggest that altL is essential for the growth of RAG-1 on C28.

FIG 1.

Identification of potential alkane transporters screened by transcriptome analysis in Acinetobacter venetianus RAG-1. (A and B) Growth curves (A) and degradation rates (B) of the wild-type strain (WT) and ΔF959_RS02655, ΔF959_RS06225, ΔF959_RS06795, ΔF959_RS14525 (named ΔaltL), ΔaltL/P (ΔaltL containing pMMB67EH vector), and ΔaltL/PaltL (ΔaltL containing the altL complemented plasmid) strains cultured in C28 BSM medium for 5 days. Values are shown as means ± SD (n = 3) and were analyzed by independent-sample t test with the WT control (**, P < 0.01; ns, not significant).

BOCTOPUS 2 prediction showed that AltL has a typical structure with transmembrane β-strands forming a transmembrane β-barrel protein (Fig. S2A). Signal peptide prediction revealed a 25-amino acid signal peptide sequence at the N terminus (Fig. S2B). Conserved domain analysis showed that AltL belongs to the OMPP1/FadL/TodX (FadL) family (Fig. S2C) and shares 20%, 21%, 22%, and 25% sequence identity with FadL family transporters FadL_Ec (26), TbuX (19), TodX (18), and XylN (27), respectively. This indicated that AltL in the OM might be responsible for the transport of long-chain alkane (C₂₈) across the OM.

The alkRa-gfp system can be used for intracellular alkane detection.

To explore the function of AltL in alkane transport, an intracellular alkane detection system was established (Fig. 2A). Alkane hydroxylase (AH) AlkMa in RAG-1 is responsible for oxidation of alkanes with a chain length ranging from C₁₆ to C₂₈ (28). Transcriptome analysis showed that expression of alkMa (F959_RS09060) and AraC family transcriptional regulator alkRa (F959_RS09055), adjacent to alkMa in the genome, was upregulated 152.30- and 3.02-fold, respectively, when cells were cultured with C₂₈ compared with SA (Table S2). This indicates that alkMa might be induced by C₂₈ or its metabolites produced during alkane metabolism in RAG-1. In Acinetobacter sp. strain ADP1, inducible transcription of AH gene alkM is activated by the regulator AlkR (29). BLASTP analysis showed that AlkRa and AlkMa share 63% and 84% sequence identity with AlkR and AlkM, respectively, in ADP1 (Table S2), suggesting that expression of alkMa may also rely on activation of AlkRa. To test this hypothesis, alkRa deletion mutants (ΔalkRa) were constructed, and transcription of alkMa and alkRa was measured. Deletion of alkRa downregulated the expression of alkMa, and the complemented strain (ΔalkRa/PalkRa) partially restored expression of alkMa (Fig. 2B), indicating that alkRa is necessary for induced expression of alkMa.

FIG 2.

Construction of the intracellular alkane detection system. (A) Schematic diagram of the construction process. alkMa and almA are alkane hydroxylases responsible for C₂₈ oxidation in Acinetobacter venetianus RAG-1; alkRa is a putative transcriptional activator that is essential for the expression of alkMa; altL is the potential alkane transporter; gfp represents the green fluorescent protein gene, which replaced the alkMa gene in ΔalmA/alkMa::gfp-pKRG and ΔaltL/almA/alkMa::gfp-pKRG as a reporter. almA was also deleted in ΔalmA/alkMa::gfp-pKRG and ΔaltL/almA/alkMa::gfp-pKRG to eliminate the process of C₂₈ degradation in RAG-1. The expression of gfp can be activated in the presence of alkRa and C₂₈; hence, the expression of GFP can be used as a reporter to reflect the presence of intracellular alkanes in ΔalmA/alkMa::gfp-pKRG. (B) Transcriptional levels of genes alkRa and alkMa in strains WT, ΔalkRa, and ΔalkRa/PalkRa with C₂₈ as the sole carbon source. The expression of alkRa and alkMa in the WT strain was used as a control. (C) GFP fluorescence in strain ΔalmA/alkMa::gfp-pKRG in the absence or presence of C₂₈. GFP fluorescence was calculated as relative fluorescence units (F/OD), which represent fluorescence values per OD₆₀₀. SA represents BSM medium supplemented with 0.5% sodium acetate, and SA+C28 represents BSM medium supplemented with 0.5% sodium acetate and 0.1% C₂₈. Values are shown as means ± SD (n = 3) and were analyzed by independent-sample t test (**, P < 0.01; ***, P < 0.001).

To explore whether the alkRa-alkMa system was induced by C₂₈ rather than its metabolites, a C₂₈ utilization-deficient strain was constructed by deleting AH genes alkMa and almA, which are responsible for C₂₈ oxidation (28), and the green fluorescent protein (GFP) gene was introduced as a reporter to yield strain ΔalmA/alkMa::gfp (Fig. 2A). Expression of gfp in ΔalmA/alkMa::gfp could be induced in the presence of C₂₈ even if the strain could not metabolize C₂₈ (Fig. S3A), indicating that alkRa-gfp could be directly induced by C₂₈. However, GFP fluorescence in the presence of C₂₈ was not detected until the fifth day (Fig. S3B), which might be due to low expression levels of GFP in ΔalmA/alkMa::gfp. To overcome this problem, The ΔalmA/alkMa::gfp strain was transformed with alkRa-gfp using the recombinant plasmid pKRG, generating alkRa-gfp overexpression strain ΔalmA/alkMa::gfp-pKRG. Transcriptional expression of gfp and GFP fluorescence were measured at different times in cells cultured in SA medium and SA medium supplemented with C₂₈ (SA+C₂₈). The overexpression strain displayed significantly increased gfp expression and GFP fluorescence in the presence of C₂₈, and relatively stable expression was detected after 6 h of incubation (Fig. 2C and Fig. S4). These results indicated that the system was suitable for the detection of intracellular long-chain alkanes.

AltL is responsible for C₂₈ transport in RAG-1.

To explore the function of AltL, the altL gene was deleted in ΔalmA/alkMa::gfp-pKRG, generating strain ΔaltL/almA/alkMa::gfp-pKRG. Strains ΔalmA/alkMa::gfp-pKRG and ΔaltL/almA/alkMa::gfp-pKRG were cultured in SA+C₂₈ medium, and expression of gfp and GFP fluorescence were measured. Deletion of altL downregulated gfp expression in the presence of C₂₈ compared with expression in strain ΔalmA/alkMa::gfp-pKRG (Fig. 3A). GFP fluorescence was also decreased in the altL deletion mutant, and the fluorescence value per the optical density at 600 nm (OD₆₀₀) (F/OD) for strain ΔaltL/almA/alkMa::gfp-pKRG in the presence of C₂₈ was similar to that in the absence of C₂₈ (Fig. 3B). This indicates that deletion of altL resulted in the decrease or absence of C₂₈ in cells, even when cells were growing vigorously (data not shown).

FIG 3.

AltL is responsible for C28 transport. (A) Induced expression of gfp gene in strains ΔalmA/alkMa::gfp-pKRG and ΔaltL/almA/alkMa::gfp-pKRG in the presence of C₂₈. The expression levels of gfp in ΔalmA/alkMa::gfp-pKRG was normalized to 1. (B) GFP fluorescence in strains ΔalmA/alkMa::gfp-pKRG and ΔaltL/almA/alkMa::gfp-pKRG in the presence of C₂₈. GFP fluorescence was calculated as relative fluorescence units (F/OD), which represent fluorescence values per OD₆₀₀. Values are shown as means ± SD (n = 3) and were analyzed by independent-sample t test (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, not significant). (C) Growth curves of strains WT, ΔaltL, and ΔaltL/PaltL cultured in SA medium supplemented with C₂₈ at 30°C. The inset shows the growth of strains in the initial 3 h of culture. (D) Changes in δ¹³C_V-PDB value for strains WT, ΔaltL, and △altL/PaltL cultured in SA medium supplemented with C₂₈-1,2-¹³C₂ at 30°C measured at different times. The inset shows the changes of δ¹³C_V-PDB value in the initial 8 h of culture. SA represents BSM medium supplemented with 0.5% sodium acetate, and SA+C28 represents BSM medium supplemented with 0.5% sodium acetate and 0.1% C₂₈. Values are from one independent experiment.

To further confirm whether C₂₈ is completely absent in cells lacking altL, transport assays were performed using carbon isotope-labeled C₂₈-1,2-¹³C₂ as the substrate. Growth curves and intracellular isotopic carbon (¹³C) levels (represented by δ¹³C_V-PDB) were measured for the WT, ΔaltL, and ΔaltL/PaltL strains in SA medium supplemented C₂₈-1,2-¹³C₂. The results showed that all three strains grew rapidly in the medium from the beginning of the culture (Fig. 3C), and due to the exhaustion of SA, ΔaltL and ΔaltL/PaltL eventually exhibited decreased growth (Fig. S5). Consistent with cell growth, δ¹³C_V-PDB levels increased in WT and ΔaltL/PaltL cells throughout culture (Fig. 3D), indicating that C₂₈-1,2-¹³C₂ was utilized by these strains expressing altL. However, C₂₈-1,2-¹³C₂ and its metabolites were almost absent in ΔaltL cells (Fig. 3D). These results indicate that deletion of altL completely abolishes the transport of C₂₈ into cells; hence, AltL is an essential protein for C₂₈ transport.

AltL mediates the transport of broad-spectrum n-alkanes and (very) long-chain fatty acids.

Due to the similar characteristics of alkanes, we wondered whether other n-alkanes were also transported by AltL. Therefore, n-alkanes of differing chain lengths, including C₁₆, C₁₈, C₂₀, C₂₄, C₃₂, and C₃₈, were tested as sole carbon sources, and cell growth was measured in the presence and absence of the altL gene. The ΔaltL mutant could not grow on n-alkanes C₂₀, C₂₄, C₃₂, and C₃₈ and exhibited reduced growth on C₁₆ and C₁₈ (Fig. 4A to E and S6). These results indicate that AltL is essential for long-chain alkane transport (C₂₀ to C₃₈) and plays an important role in the transport of medium-chain alkanes (C₁₆ to C₁₈). To determine whether AltL also plays a role in fatty acid transport like other FadL family members, fatty acids of differing chain lengths, including palmitic acid (C_16A), stearic acid (C_18A), arachidic acid (C_20A), n-tetracosanoic acid (C_24A), and n-octacosanoic acid (C_28A) were tested as carbon sources in the cultivation of WT, ΔaltL, and ΔaltL/PaltL strains. ΔaltL showed no significant growth difference with WT under C_16A and C_18A (Fig. 4F and G) but almost completely lost the growth ability on C_20A, C_24A, and C_28A as sole carbon sources (Fig. 4H to J), suggesting that AltL also mediates the transport of very long-chain fatty acids (C_20A to C_28A).

FIG 4.

Effects of AltL on alkane and fatty acid utilization. (A to E) Growth curves of strains WT, ΔaltL, and ΔaltL/PaltL cultured in BSM medium with C₁₆, C₁₈, C₂₀, C₂₄, and C₃₂ as the sole carbon source. (F to J) Growth curves of strains WT, ΔaltL, and ΔaltL/PaltL cultured in BSM medium with C_16A, C_18A, C_20A, C_24A, and C_28A as the sole carbon source.

Previous studies showed that FadL family member FadL_Ec can promote the uptake of C₈ and C₁₀ n-alkanes (20). To identify other alkane transporters in RAG-1, genome analysis was performed using BLASTP with FadL_Ec, and a similar protein, FadL_Ra (F959_RS04450), was found to share 22% identity with FadL_Ec. We next generated ΔfadL_Ra and ΔaltL/fadL_Ra mutants by insertional inactivation, and the growth of WT, ΔfadL_Ra, and ΔaltL/fadL_Ra strains on C₁₆, C₁₈, and C₂₀ was measured. Insertional inactivation of fadL_Ra led to reduced growth on C₁₆ and C₁₈ but had no significant impact on C₂₀ (Fig. S7A to C). Double-gene disruption of altL and fadL_Ra decreased growth on C₁₆, C₁₈, and C₂₀ (Fig. S7A to C). These results indicate that in addition to AltL, FadL_Ra can also mediate the transport of C₁₆ and C₁₈ alkanes. In addition, the abilities of ΔfadL_Ra and ΔaltL/fadL_Ra to utilize C_16A, C_18A, and C_20A were also determined. Growth of the two mutants was not affected on C_16A (Fig. S7D), but double-gene deletion decreased the growth of ΔaltL/fadL_Ra on C_18A (Fig. S7E), suggesting that AltL and FadL_Ra both mediate the transport of C_18A.

Longer extracellular loops may support the transport of more hydrophobic and larger substrates.

Compared with other known FadL family members (18, 19, 26), AltL (555 amino acids) is ~100 residues longer (Table 1). AltL also shares low sequence identity and coverage with other members and preferentially transports more hydrophobic and larger substrates (Table 1). Moreover, AltL has no similarity with the reported long-chain alkane transporters OmpT-1 and OmpT-3, and only 23% identity was found with OmpT-2 at 116-amino acid coverage (14). These characteristics imply that AltL may have a structure different from other FadL family members and reported alkane transporters.

TABLE 1.

Comparison of AltL and other FadL family members

Protein	Length (aa)a	Sequence identity [n (%)]b	Substrate(s)	Molecular weightc	Strain	Reference
AltL	555		C16−C38	535.03	RAG-1
			C18A−C28A	424.74
FadLEc	446	44/217 (20)	C8−C10	142.28	E. coli	21
			C7A−C18:1A	282.46
TbuX	458	80/380 (21)	Toluene	92.14	R. pickettii PKO1	19
TodX	453	69/319 (22)	Toluene	92.14	P. putida F1	18
CymD	460	70/321 (22)	Toluene	92.14	P. putida F1	66
XylN	465	103/417 (25)	m-Xylene	106.16	P. putida	27
AupA	452	26/76 (34)	C16	226.44	M. hydrocarbonoclasticus SP17	13
OmpT-1	376	0	C28–C36, pristane	506.97	A. dieselolei B5	14
OmpT-2	443	27/116(23)	C16–C24	338.65	A. dieselolei B5	14
OmpT-3	449	0	C8–C12	170.33	A. dieselolei B5	14

^aaa, amino acids.

^bIdentities were calculated using BLASTP.

^cMolecular weight refers to the maximum molecular weight of the substrate that the protein can transport.

To explore the structural features of AltL, a homology model was built based on the structure of FadL_Pa (PDB 3DWO) from Pseudomonas aeruginosa (20.42% identity) (30) and optimized by molecular dynamics (MD) simulation. The model has a β-barrel structure composed of 14 antiparallel β-strands, and the interior was blocked by an N-terminal hatch domain (Fig. 5A). Similar to other FadL family members (FadL_Ec, TodX, and TbuX), an inward-pointing kink is present in strand S3, which forms a lateral opening in the β-barrel wall (Fig. 5A and B). The main difference between AltL and other FadL family members is the longer extracellular loops in AltL than the corresponding sequences in FadL_Ec, TodX, and TbuX, specifically for loops L1, L3, L4, L5, and L7 (Fig. 5B and Table S3).

FIG 5.

Structural features of AltL. (A) Ribbon diagram of the final stable structure of AltL from homology modeling and molecular dynamics simulation. The approximate position of the outer membrane (OM) is indicated by two horizontal lines. E and P represent extracellular and periplasmic sides, respectively. (B) Comparison of the AltL homology model with the structures of FadL family members FadL_Ec and TodX, shown in ribbon representations. Extracellular loops are colored cyan (L3), green (L4), yellow (L5), and magenta (L7). The S3 strand is colored blue. Bound substrate molecules are colored red. (C) Molecular docking between substrate C₂₈ and AltL. Substrate molecules are colored cyan. (D) Molecular docking between substrate C₁₆ and AltL. Substrate molecules are colored cyan.

To explore the relationship between the structure of the extracellular loops and the binding and transport of substrates, molecular docking was performed with long-chain alkane C₂₈ and medium-chain alkane C₁₆. The binding model between AltL and C₂₈/C₁₆ near extracellular loops showed that AltL and substrates exhibited steric complementarity with binding energies of −7.9 and 7.6 kcal/mol, respectively. Van der Waals (VDW) interactions were observed between substrates and surrounding amino acid residues. C₂₈ is mainly enclosed by residues of loop L4 and some residues of loops L5 and L7 (Fig. 5C and Table S4). C₁₆ is also mainly buried by residues of loop L4 and some residues of loops L3, L5, and L7 (Fig. 5D and Table S5). Loop L4 forms the most intimate interactions and might play a key role in substrate binding, but residues of L3, L5, and L7 may also contribute. The longer extracellular loops compared with other family members may support the binding of larger, more hydrophobic substrates.

AltL homologs are widespread in Gamma-, Beta-, and Deltaproteobacteria.

In order to probe the ecophysiological role of the novel long-chain alkane transporter, a phylogenetic tree was constructed using AltL homologs with sequence identity of >30% (Fig. S8). Homologs of AltL are widely distributed in some orders of the Gamma-, Beta-, and Deltaproteobacteria classes (Fig. S8), among which Gammaproteobacteria contains the most homolog sequences. Six genera, Acinetobacter and Pseudomonas (order Pseudomonadales), Alcanivorax, Ketobacter, and Oleiphilus (order Oceanospirillales), and Marinobacter (order Alteromonadales), possess AltL homologs (Fig. S8), and these organisms include many important hydrocarbon degraders present in marine and terrestrial environments (31–33). Sequence analysis showed that they share 72 to 100%, 39 to 44%, 48 to 51%, 27 to 36%, 34 to 41%, and 40 to 43% identity with AltL in RAG-1, respectively. Other genera in the Pseudomonadales, Oceanospirillales, and Alteromonadales orders also contain AltL homologs (Fig. S8). In the class Betaproteobacteria, AltL homologs are only found in certain genera of the order Burkholderiales, and they share 30 to 37% identity with AltL (Fig. S8). In the class Deltaproteobacteria, AltL homologs are distributed in the genera Desulfatibacillum, Desulfobacterales, Desulfococcus, Desulfoluna, Desulfosarcina, and Desulfatitalea (order Desulfobacterales) and Smithella (order Syntrophobacterales), sharing ~30% identity to AltL (Fig. S8).

The hydrocarbon-degrading strains reported among these genera possessing AltL homologs are summarized in Table S6. Most genera containing AltL homologs also contain strains capable of degrading alkanes, especially in the Acinetobacter, Pseudomonas, Marinobacter, and Alcanivorax genera. Long-chain alkane-degrading strains have also been reported in the order Burkholderiales (Betaproteobacteria) and families Desulfobacteraceae and Syntrophaceae (Deltaproteobacteria) (34–37). These results suggest a positive correlation between AltL homologs and the alkane degradation ability of strains.

Phylogenetic analysis showed that most of these homologous genes exist in specific gene clusters (alt gene clusters), containing three conserved genes encoding a C39 family peptidase, a hypothetical protein, and an outer membrane protein, along with some other variable genes (Fig. 6). The same organization and sequence conservation of the first three genes in alt gene clusters implies that these genes may perform specific functions, such as alkane transport. In addition, homologs of altL in Ketobacter were not identified in the above-mentioned gene clusters (Fig. 6), suggesting a different evolutionary course from other genera.

FIG 6.

Organization of gene clusters carrying AltL homologs. Possible genes in these clusters are surrounded by bold black boxes, and homologous genes are shown in the same colors. Three conserved genes are present in these gene clusters, including genes colored cyan, yellow, and pink, except genes in Ketobacter.

An AltL homolog in P. aeruginosa also functions in alkane transport.

To verify the function of AltL homologs in other genera, we tested Pseudomonas aeruginosa PAO1 in the genus Pseudomonas, a typical alkane-degrading strain (38, 39). A similar alt gene cluster was found in PAO1, containing four genes. The first three conserved proteins share 54.04, 42.31, and 39.64% identity with homologs in RAG-1 (Fig. 7A). The altL homologous gene PA1764 (named altL_Pa) in PAO1 was deleted, and the growth of the mutant was measured in media containing different chain-length n-alkanes and fatty acids. Deletion of altL_Pa resulted in the growth defect of strains on n-alkanes C₁₆, C₂₀, C₂₄, C₂₈, and C₃₂ (Fig. 7B and Fig. S9) but had no significant effect on fatty acids C_12A, C_16A, C_18A, and C_20A, although growth was diminished on C_24A (Fig. S10). These results indicate that the AltL homolog in P. aeruginosa performs a similar function in alkane transport.

FIG 7.

The functions of AltL homologs in Pseudomonas aeruginosa. (A) Comparison of the alt gene cluster in RAG-1 and PAO1. Homologous proteins are shown in the same color. (B) Growth curves of PAO1 (WT), ΔaltL_Pa, and ΔaltL_Pa/PaltL_Pa cultured with different chain-length n-alkanes (C₁₆, C₂₀, C₂₄, C₂₈, and C₃₂) as the sole carbon source.

Heterologous expression of AltL enhances alkane degradation in Pseudomonas aeruginosa.

To explore the potential for the application of AltL in engineered alkane-degrading hosts, an AltL heterologous expression strain (PAO1/PaltL) was constructed, and its growth and degradation abilities were compared with those of the PAO1/P control strain. Heterologous expression of AltL from RAG-1 increased cell growth for all tested n-alkanes, including C₁₆, C₂₀, C₂₄, C₂₈, and C₃₂ with increased degradation rates from 52.01%, 59.99%, 34.42%, 19.63%, and 13.92% to 61.38%, 68.33%, 41.37%, 31.20%, and 18.93%, respectively (Fig. S11), consistent with the transport role of AltL on these n-alkanes. These results suggest that alkane transport might be a limiting step for alkane degradation, and the alkane transporter AltL has great potential in the construction of engineered strains for oil pollution bioremediation, and industrial production.

DISCUSSION

Alkane uptake is the first step of alkane biodegradation. The detection of intracellular alkanes is an important part of the study of alkane uptake. In this study, we constructed an alkane detection system based on the alkRa-alkMa operon induced by alkanes to detect the presence of alkanes in cells. alkRa-gfp obtained by replacing alkMa with gfp can only reflect the existence of intracellular alkanes to a certain extent but cannot quantify the content of alkanes in cells. In a previous study, the ddvR-lacZ system induced by 5,5′-dehydrivanillate (DDVA) was used to detect the DDVA uptake ability of outer membrane protein DdvK, which was mainly reflected by the β-galactosidase activity (40) and showed the sensitivity of the system to intracellular DDVA. Currently, the specific mechanism of action between alkRa and alkMa is still unclear. The AlkRa homolog AlkR in Acinetobacter sp. strain ADP1 has been shown to be necessary for the induced expression of alkM as a transcriptional activator, and alkanes with a chain length from heptane to octadecane can induce the expression of alkM (29), which reflects the relationship between alkanes and the alkR-alkM operon. It can also be observed in the regulation mechanism of other alkane hydroxylase expressions (41, 42). AlkS in Pseudomonas oleovorans GPo1 is located upstream of the alkane degradation gene cluster, which activates the expression of the alkBFGHJKL operon in the presence of alkanes (41, 43). In species of the Gram-positive alkane-degrading bacterium Dietzia, the expression of alkane hydroxylase CYP153 was induced by n-alkanes composed of 8 to 14 carbon atoms but not by derived decanol and decanoic acid (42). In our study, we demonstrated that alkanes themselves can induce the expression of the alkRa-gfp system by knocking out the starting alkane hydroxylase genes to block the downstream metabolic pathway (Fig. 2C and D). However, it is still unclear how alkanes interact with alkRa and alkMa to activate the expression of alkMa.

To detect the alkane content in cells, an isotope transport experiment was conducted. Due to the growth defect of the ΔaltL mutant, a small amount of SA (0.5 g/L) was added as a supplementary carbon source to support the growth of bacteria. In this process, we found that the altL complemented strain growth similar to that of the deletion mutant at the early stage of culture (38 h), and then its growth surpassed the deletion mutant after 38 h (Fig. 3C). The reason why the growth of complemented strain in the medium supplemented with SA is weaker than that with a single alkane may be that the rapid growth of cells in the early culture with a preferred carbon source of SA led to plasmid loss in some cells along with the degradation of ampicillin (44, 45), which led to the failure of some strains to quickly switch to the mode of utilizing alkanes after the absence of the preferred carbon source in the later period. Compared with the growth curve, the accumulation of intracellular ¹³C isotopes has some puzzling aspects, which are mainly reflected in the poor growth of the complemented strain in the early stage of the growth curve (Fig. 3C), while the continuous accumulation of ¹³C isotopes in the cell was detected at the 4th hour in the accumulation curve (Fig. 3D). We speculate that the preferred carbon sources resulted in the unobvious growth difference between the deletion strain and the complemented strain, while due to the existence of alkane transporter AltL in the complemented strain, it can still absorb some alkanes and accumulate in the cells. However, our current research cannot provide direct evidence of this hypothesis.

Long-chain fatty acids and long-chain alkanes are highly hydrophobic and have low water solubility. In RAG-1, AltL simultaneously mediates the transport of long-chain alkanes and (very) long-chain fatty acids, similar to FadL_Ec, which is capable of transporting both long-chain fatty acids (C_7A to C_18:1A) and short-chain alkanes (C₈ to C₁₀) (20, 46). This phenomenon reflects substrate compatibility of the FadL family transporters to substrates with similar structures and hydrophobicities.

In FadL_Ec, the first interaction with the substrate occurs in a hydrophobic groove between loops L3 and L4 (21), whereas in TodX and TbuX, a hydrophobic cleft for substrate binding is formed by loop L3 alone (22). Compared with FadL_Ec, TodX, and TbuX, AltL has longer extracellular loops (Fig. 5B and Table S3). Molecular docking between AltL and C₂₈/C₁₆ showed more loops involved in alkane binding, including L3, L4, L5, and L7, all of which are longer than in other family members. These multiple long extracellular loops may help AltL bind more hydrophobic and larger substrates. Loops are the most flexible regions of β-barrel proteins and are very important for their functions (47), and they can adopt multiple conformations to facilitate substrate binding (48, 49). A dynamic translocation pathway in short- and medium-chain alkane transporter AlkL has been revealed in which extracellular loops form a barrel extension and enlarge the narrow lateral opening of AlkL to accommodate substrates (12). In Sinorhizobium meliloti, predicted extracellular loop L5 in FadL_Sm and other α-rhizobial FadL proteins contain specificity determinants for long-chain N-acyl homoserine lactones (AHLs) (50). These extracellular loops in AltL that differ from those in other FadL proteins might be the key features that recognize and bind large, highly hydrophobic substrates.

AltL homologs are widespread in Gamma-, Beta-, and Deltaproteobacteria. In class Gammaproteobacteria, TOL1625 in Thalassolituus oleivorans MIL-1 has been linked to long-chain alkane (C₂₈) transport based on transcriptome analysis (51). A BLAST search revealed 39.68% identity and 96% coverage between TOL1625 and AltL, which provides further evidence that AltL homologs may be involved in the transport of long-chain alkanes in marine obligate hydrocarbonoclastic bacteria (OHCB).

In class Deltaproteobacteria, AltL homologs are found in various genera of the orders Desulfobacterales and Syntrophobacterales. Some isolates from these two orders have the ability to anaerobically degrade medium- and long-chain n-alkanes, such as Desulfococcus oleovorans strain Hxd3 (C₁₂ to C₂₀), Desulfatibacillum alkenivorans strain AK-01 (C₁₃ to C₁₈), and Desulfatibacillum strain Pnd3 (C₁₄ to C₁₇) (52) in order Desulfobacterales, and Smithella spp. (37) in the order Syntrophobacterales. This leads us to speculate that anaerobic bacteria may also rely on AltL homologs for alkane transport, although they have different alkane degradation pathways from aerobic bacteria. This type of transport mechanism based on diffusion may be ideal for both aerobic and anaerobic bacteria to obtain hydrophobic substances from the environment.

In summary, we identified AltL, a novel transporter responsible for the transport of strongly hydrophobic substrates in RAG-1. Our results highlight a remarkable distinction in substrates and extracellular loop structure compared with other FadL members and a widespread distribution in bacteria. AltL could be applied to engineer alkane-degrading bacteria for oil pollution bioremediation and modified to improve substrate compatibility for transporting other strongly hydrophobic substrates. Although the transport mechanism of AltL for strongly hydrophobic substrates is still unclear, its unique structural features provide insight into the selectivity of the transporter for hydrophobic substrates.

MATERIALS AND METHODS

Bacterial strains and culture conditions.

The bacterial strains, plasmids, and primers used in this study are listed in Table 2 and Table 3. Bacteria were cultured in Luria-Bertani medium (LB; 10 g/L tryptone, 5 g/L yeast extract, 5 g/L NaCl, pH 7.0) or basal salt medium [BSM; 3.815 g/L K₂HPO₄, 0.5 g/L KH₂PO₄, 0.825 g/L (NH₄)₂HPO₄, 1.2625 g/L KNO₃, 0.2 g/L Na₂SO₄, 0.02 g/L MgCl₂, 0.02 g/L CaCl₂, 0.002 g/L FeCl₃, pH 7.2] (53) supplemented with different carbon sources at 30°C with shaking at 200 rpm. SA was added at 0.2% (wt/vol); n-alkanes were added at 0.1% (vol/vol) for C₁₆ and 0.1% (wt/vol) for C₁₈ to C₃₈; (very) long-chain fatty acids were added at 0.1% (wt/vol) for C_12A to C_20A and 0.05% (wt/vol) for C_24A to C_28A. When necessary, antibiotics were added at the following concentrations (μg/mL): kanamycin, 50; chloramphenicol, 10; ampicillin, 150 for RAG-1 and its derivatives; tetracycline, 50; and kanamycin, 25 for PAO1 and its derivatives. Isopropyl β-d-1-thiogalactopyranoside (IPTG) was added to the medium at an appropriate concentration to induce expression when needed.

TABLE 2.

Strains and plasmids used in this study

Strain/plasmid	Characteristicsa	Source
Acinetobacter venetianus strains
WT	Acinetobacter venetianus RAG-1, WT strain	Presented by Professor D L Gutnick
ΔF959_RS02655	F959_RS02655 deletion mutant	This study
ΔF959_RS06225	F959_RS06225 deletion mutant	This study
ΔF959_RS06795	F959_RS06795 deletion mutant	This study
ΔaltL	F959_RS14525 (altL) deletion mutant	This study
ΔaltL/PaltL	ΔaltL complemented strain	This study
ΔalkRa	alkRa deletion mutant	This study
ΔalkRa/PalkRa	ΔalkRa complemented strain	This study
ΔalmA/alkMa::gfp	Mutant with almA deletion and replacement of alkMa by gfp	This study
ΔaltL/almA/alkMa::gfp	Mutant with almA and altL deletion and replacement of alkMa by gfp	This study
ΔalmA/alkMa::gfp-pKRG	Mutant ΔalmA/alkMa::gfp containing pKRG plasmid	This study
ΔaltL/almA/alkMa::gfp-pKRG	Mutant ΔaltL/almA/alkMa::gfp-pKRG containing pKRG plasmid	This study
ΔfadLRa	fadLRa insertion inactivation mutant	This study
ΔaltL/fadLRa	altL deletion and fadLRa insertion inactivation mutant	This study
Pseudomonas aeruginosa strains
PAO1	WT strain	Lab collection
ΔaltLPa	PA1764 deletion mutant	This study
ΔaltLPa/PaltLPa	ΔaltLPa complemented strain	This study
PAO1/P	PAO1 containing pMMB67EH plasmid	This study
PAO1/PaltL	PAO1 containing PaltL plasmid	This study
Escherichia coli strains
DH5α	Cloning strain	Lab collection
S17-1	recA pro hsdR RP4-2-Tc::Mu-Km::Tn7	Lab collection
Plasmids
pK18mobsacB	Suicide plasmid for gene knockout, Kanr	Lab collection
pKU	Suicide plasmid for gene knockout, Kanr	Lab collection
pMMB67EH	Expression vector, Ampr	Lab collection
pK18-F959_RS02655-UD	pK18mobsacB containing the 904-bp upstream fragment and 871-bp downstream fragment of F959_RS02655 gene, Kanr	This study
pK18-F959_RS06225-UD	pK18mobsacB containing the 787-bp upstream fragment and 728-bp downstream fragment of F959_RS06225 gene, Kanr	This study
pK18-F959_RS06795-UD	pK18mobsacB containing the 796-bp upstream fragment and 709-bp downstream fragment of F959_RS06795 gene, Kanr	This study
pK18-altL-UD	pK18mobsacB containing the 697-bp upstream fragment and 743-bp downstream fragment of altL gene, Kanr	This study
pK18-almA-UD	pK18mobsacB containing the 793-bp upstream fragment and 719-bp downstream fragment of almA gene, Kanr	This study
pK18-alkMa-UD	pK18mobsacB containing the 830-bp upstream fragment and 635-bp downstream fragment of alkMa gene, Kanr	This study
pK18-gfp-UD	pK18mobsacB containing the 830-bp upstream fragment of alkMa, gfp gene, and 635-bp downstream fragment of alkMa, Kanr	This study
pMK	pMMB67EH with replacement of Ampr to Kanr and deletion of lacI	This study
pKRG	pMK carrying the alkRa-gfp sequence	This study
PaltL	pMMB67EH carrying the altL ORF	This study
PalkRa	pMMB67EH carrying the alkRa ORF	This study
PaltLPa	pMMB67EH carrying the altLPa ORF	This study

^aORF, open reading frame.

TABLE 3.

Primers used in this study

Target gene	Primer name	Sequence (5′−3′)a
Primers for qRT-PCR
16S rRNA	16S-RTF	ACAGAGGGTGCGAGCGTTAATC
	16S-RTR	CTGCCTTCGCCATCGGTATTCC
F959_RS02655	F959_RS02655-RTF	AGCATTCTTAGCACCGAACACAACA
	F959_RS02655-RTR	CTCCCAATTTGCGGCACCTGAAA
F959_RS06225	F959_RS06225-RTF	GCAGTAGGTTGGCTTGCAGGTT
	F959_RS06225-RTR	TTGTGGCGTCGTGATTGTGGTT
F959_RS06795	F959_RS06795-RTF	CATGCCATGCCACAAGGAAGTG
	F959_RS06795-RTR	ACTGTTGCCGTGCCTGTGAG
altL	altL-RTF	ACTGTGGGAGCCAACTGATGACTTT
	altL-RTR	GCAGCAAGAATCTGACCCGTAGCA
alkRa	alkRa-RTF	CCGCTATCCACAACCATCCTGA
	alkRa-RTR	TTGGCTCGCTAAACGCATTCG
alkMa	alkMa-RTF	CGATGGGTGCGATCAACGGTAT
	alkMa-RTR	TGGTGTTGCAGCACGTTTATGG
gfp	gfp-RTF	CGGTGAAGGTGAAGGTGATGC
	gfp-RTR	CTTCTGGCATGGCAGACTTGAA
Primers for gene deletion
F959_RS02655	F959_RS02655-F1	CGGGATCCAACTAACCCAAACGGCTTCTA
	F959_RS02655-R1	CACTTGTGGAATGGTCGAAAACCATGTCTCGAATCAGTTTATC
	F959_RS02655-F2	GATAAACTGATTCGAGACATGGTTTTCGACCATTCCACAAGTG
	F959_RS02655-R2	CCCAAGCTTATGCGGGTTAGCAGAAGTGA
F959_RS06225	F959_RS06225-F1	CGGGATCCAGGTCTAATTTGCTTGAACCTGC
	F959_RS06225-R1	CAAGCCATAGAGGCATCATCCTTGTTAAACC
	F959_RS06225-F2	GATGATGCCTCTATGGCTTGAAAATCGGCTAT
	F959_RS06225-R2	AACTGCAGAATAACGATACGGTATAAATAGCCC
F959_RS06795	F959_RS06795-F1	CGGGATCCTCCAACACTCTATTTGTACTTCGG
	F959_RS06795-R1	TCGATATCAAGATGGGGGACTCCATGTCC
	F959_RS06795-F2	GTCCCCCATCTTGATATCGACCCACTGATTACA
	F959_RS06795-R2	AACTGCAGTATCGCATAGTGGACCGACA
altL	altL-F1	CGGAATTCTACTGTTCAATATGGCGTCTTAG
	altL-R1	TCTTTTGCATCATGTCTCCTTAACGTAGTTTTTTT
	altL-F2	AGGAGACATGATGCAAAAGAACAGAAATCAACG
	altL-R2	GCTCTAGAGCTGATAAACGCATGTAACTTTC
alkRa	alkRa-F1	CGGGATCCCATAGAATGTTTCACCCATTTGT
	alkRa-R1	TGTTTAGGTGATGATTTTTTATCGACGTGTTCA
	alkRa-F2	AAAAAATCATCACCTAAACAATATCGGCAGC
	alkRa-R2	AACTGCAGGGTGAAGAGACGTGGTTAGAAAT
almA	almA-F1	CGGGATCCTTCTGTAAATTCAAGATAGTGCCT
	almA-R1	CAAGCAAAAAATACCCTGTATAAACCATGACTTT
	almA-F2	TACAGGGTATTTTTTGCTTGAATAATTTGAAAC
	almA-R2	AACTGCAGGTACCAATGAGATCATGAAAGAACT
alkMa	alkMa-F1	CGGGATCCTGTGGATAGCGGCTAAACTA
	alkMa-R1	CTACTTGAGTCATCTTTTCCATGTCTTTGTATA
	alkMa-F2	GGAAAAGATGACTCAAGTAGAAGTAGTCGCTGA
	alkMa-R2	AACTGCAGTTCTATATCAGGTAATGAGGGCT
altLPa	altLPa-F1	GGAATTCGCGGCGCTGATGCTGTCT
	altLPa-R1	CAGCCCTCAGAAGGTCGCGGGTTACTGGGCCTTC
	altLPa-F2	GCCCAGTAACCCGCGACCTTCTGAGGGCTGGAGC
	altLPa-R2	CAAGCTTCGATGAAAGCGTAGCGATACC
Sequencing primers for gene deletion mutants
F959_RS02655	seq_F	GCACCACCAATGAAGGGG
	seq_R	GCACCATCAACAACTAAAGCC
F959_RS06225	seq_F	TCAGCATTGCAGCTAAAGTAA
	seq_R	CGCTATTCATCTATGGTGGC
F959_RS06795	seq_F	GATACAATCTGTCGCAACCA
	seq_R	CCAATCAATAAGCGGAAAA
altL	seq_F	CGGTTAAGGGTTGAAGAGG
	seq_R	TCAAGAAAGCAACGGTCAT
alkRa	seq_F	GCAACAATTCATTGTCTGTAGACC
	seq_R	ATTACGTTCGCATGTTGTAGTTG
almA	seq_F	AAGTTTAGGCGTTGATGCAG
	seq_R	CCTGTTACGGCTGTAGATGC
alkMa	seq_F	GCTAAACGCATTCGATGTT
	seq_R	CAAGCCTACCTGAGACCCT
Primers for gene knock-in
gfp	alkMa-up-F	ATTCGAGCTCGGTACCCGGGTGTGGATAGCGGCTAAACTA
	alkMa-up-R	CTTTAGACATCATCTTTTCCATGTCTTTGTATA
	gfp-F	GGAAAAGATGATGTCTAAAGGTGAAGAATTATTCA
	gfp-R	TGTCTTTAAATTATTTGTACAATTCATCCATACCA
	alkMa-dn-F	GTACAAATAATTTAAAGACATGATCAAATGAAAGC
	alkMa-dn-R	GCCAAGCTTGCATGCCTGCAAACAAGCCTACCTGAGACCCT
Primers for gene overexpression
alkRa-gfp	pMK-F	GTACCCGGGGATCCTCTAGAGTC
	pMK-R	AGCCTTTTCTGAGCATGGTATTT
	alkRa-gfp-F	TACCATGCTCAGAAAAGGCTTTATGTTTCATTTTGAATATATTGC
	alkRa-gfp-R	TCTAGAGGATCCCCGGGTACTTATTTGTACAATTCATCCATACCA
Primers for gene complementation
altL	altL-F	ATTCGAGCTCGGTACCCGGGTTGGAAAAAAACTACGTTAAGG
	altL-R	CCTGCAGGTCGACTCTAGAGTTGTATAGAAACCATCTGCATATAG
alkRa	alkRa-F	CAATTTCACACAGGAAACAGATGGATGCTTTAAGTAAAATATTTG
	alkRa-R	TCTAGAGGATCCCCGGGTACTTATGTTTCATTTTGAATATATTGC
altLPa	altLPa-F	CAATTTCACACAGGAAACAGAATTCAGTAACCCGCGAGCGAGGC
	altLPa-R	TCTAGAGGATCCCCGGGTACCTCAGAAGGTCGTACGGTAGGTCAT
Sequencing primers for plasmids
pMMB67EH	pMM-F	AGCGGATAACAATTTCACACAG
	pMM-R	CGCTTCTGCGTTCTGATTTAA

^aNucleotides in bold represent restriction enzyme sites added to the 5′ region of the primers. Underlines nucleotides represent overlap sequences.

RNA extraction, quantification, and transcriptome sequencing.

RAG-1 cells were cultured with 0.1% (wt/vol) C₂₈ or 0.2% (wt/vol) SA as carbon sources and harvested at the mid-log phase. Total RNA was extracted using an RNAprep pure cell/bacterial kit (Tiangen, Beijing, China) according the manufacturer’s instructions. Total-RNA samples were prepared for Illumina next-generation sequencing using a RiboZero kit and a PrepXTM RNA sequencing (RNA-Seq) library preparation kit and sequenced on an Illumina HiSeq 2500 platform by Novogene (Tianjin, China).

qRT-PCR analysis.

qRT-PCR experiments were performed as previously described (28).

Genetic manipulations.

Gene deletion and complementation experiments in RAG-1 cells and gene deletion in PAO1 cells were conducted via homologous recombination as described previously (28, 54). Plasmids used for gene complementation or overexpression were constructed via enzyme digestion-connection or Gibson assembly. Plasmids pMMB67EH/pMK and pK18mobsacB/pKU were used as expression and homologous recombination vectors, respectively.

Growth and degradation measurement.

Strains were cultured in LB to an OD₆₀₀ of ~2.0 and then harvested by centrifugation and washed twice with BSM. Cells were diluted 1:20 in 50 mL fresh BSM with additional carbon sources and cultivated for several days. Growth curves were generated from OD₆₀₀ values for strains grown in SA and alkane medium and by bacterial counts for strains grown in (very) long-chain fatty acid medium. The remaining alkanes were measured over 5 days for RAG-1 and 7 days for PAO1 using gas chromatography (GC). Detection of alkanes and the calculation of degradation rates were carried out as described previously (28).

GFP fluorescence measurements.

Cells were grown in 50 mL SA (0.5%) medium or SA (0.5%) plus C₂₈ (0.1%) medium at 30°C with shaking at 200 rpm. Aliquots (200 μL) were used for GFP fluorescence measurement with excitation at 488 nm and emission at 520 nm (recorded as F), and 1-mL cultures were used for OD₆₀₀ measurement (recorded as OD). GFP fluorescence was calculated as relative fluorescence units (F/OD), which represent fluorescence values per OD₆₀₀. Background fluorescence of control wells containing blank medium was subtracted before calculation.

Transport assays.

Growth and transport assays were conducted using isotope-labeled C₂₈-1,2-¹³C₂. Strains RAG-1, ΔaltL, and ΔaltL/PaltL were precultured in 20 mL BSM with 0.2% (wt/vol) SA containing the appropriate antibiotics and IPTG to an OD₆₀₀ of ~0.5 to 0.6. Cells were then harvested, transferred to 20 mL BSM containing 0.025% (wt/vol) C₂₈-1,2-¹³C₂ and 0.05% (wt/vol) SA and further cultured. For growth assays, 1-mL cultures were used for OD₆₀₀ measurement at different times. For transport assays, 2-mL cell samples were collected at different times. Samples were immediately washed three times with blank BSM to remove residual C₂₈-1,2-¹³C₂ from the cell surface, freeze-dried, and analyzed by Beijing Createch Testing Technology Co., Ltd. (Beijing, China).

Homology modeling, MD simulation, and molecular docking.

Models were computed with the SWISS-MODEL server homology modeling pipeline (30), which relies on ProMod 3 (30), a comparative modeling engine based on OpenStructure 2 (55). The dimer structure of protein AltL was built and submitted to further MD simulation for optimization. All MD simulations were performed over 100 ns using Amber 16 (56).

Docking studies were carried out using the AutoDock Vina 3 program (57). Docking sites on protein targets were defined by establishing a grid box with a default grid spacing, centered on the positions of native ligands. After docking searches, the best conformation was chosen with the lowest binding energy. For more details, see methods in Text S1 in the supplemental material.

Sequence analysis.

Sequence identity searches were performed using BLAST (58), and topology prediction was conducted with BOCTOPUS 2 (59). Signal peptide prediction was performed with SignalP 5.0 (60), and conserved domain prediction and protein family classification were performed with InterPro (61).

Phylogenetic analysis.

Sequences were filtered using BLASTP (58), and redundancy was corrected with CD-HIT (62). MEGA X was used for multiple sequence alignment (63), and conserved sequences were extracted with G-blocks (64). IQ-TREE 2 was used for the phylogenetic tree construction (65). For more details, see the methods in Text S1.

Statistical analysis.

Experiments were conducted in triplicate except for one repeat of transport assays. Results are shown as means ± standard deviation (SD). The t tests were performed with Prism 8.4 (GraphPad Software, San Diego, CA); P values of ≤0.05, 0.01, and 0.001 were considered statistically significant (*), highly significant (**), and extremely significant (***), respectively.

Data availability.

The GenBank accession numbers of the AltL and PA1764 protein sequences are WP_171054784 and NP_250455, respectively.