Fejerlectin, a Lectin-like Peptide from the Skin of Fejervarya limnocharis, Inhibits HIV-1 Entry by Targeting Gp41.

Human immunodeficiency virus type 1 (HIV-1) is mainly transmitted by sexual intercourse, and effective microbicides preventing HIV-1 transmission are still required. Amphibian skin is a rich source of defense peptides with antiviral activity. Here, we characterized a lectin-like peptide, fejerlectin (RLCYMVLPCP), isolated from the skin of the frog Fejervarya limnocharis. Fejerlectin showed significant hemagglutination and d-(+)-galacturonic acid-binding activities. Furthermore, fejerlectin suppressed the early entry of HIV-1 into target cells by binding to the N-terminal heptad repeat of HIV-1 gp41 and preventing 6-HB formation and Env-mediated membrane fusion. Fejerlectin is the smallest lectin-like peptide identified to date and represents a new and promising platform for anti-HIV-1 drug development.

Introduction

Human immunodeficiency virus (HIV), which infects about 2 million people annually, remains a serious global public health problem.[1] The entry of HIV-1 into host cells relies on the membrane fusion mediated by its envelope protein (Env), which comprises the exposed surface subunit gp120 and the transmembrane subunit gp41.[2] Gp120 first binds to the receptor CD4 receptor and then interacts with the CCR5 or CXCR4 co-receptors on the host cells.[3] Subsequently, gp41 undergoes conformational changes that result in fusion of the viral envelope to the host cell membrane.[4] As the critical components in the first stages of HIV-1 infection, gp120 and gp41 have become important targets for the development of HIV-1 entry inhibitors.[5]

Lectins are a group of carbohydrate-binding proteins that recognize and reversibly bind to glycosyl ligands.[6] Gp120 and gp41 are highly glycosylated and, consequently, are important lectin targets, and many lectins have significant anti-HIV-1 activity.[7] For example, BanLec, a lectin isolated from bananas (Musa acuminata), significantly inhibits HIV-1 infection by recognizing mannose-rich structures and binding gp120.[8] Similarly, the β-galactose-specific lectin from the polychaete marine worm Chaetopterus variopedatus has also been shown to exert anti-HIV-1 activity by blocking HIV-1 entry into cells.[9] However, lectins tend to weigh over 10 kDa, resulting in unavoidable toxicity or immunogenicity, limiting their clinical development and application.[10−12] Therefore, the discovery of small, specific lectins would be highly desirable. While small-size lectins such as retrocyclin-1, Selenocosmia huwena lectin-I, and odorranalectin have been discovered, only a few have exhibited anti-HIV capacity.[13−15]

Peptides secreted by amphibian skin have been extensively studied due to their bioactivity and low cytotoxicity.[16] Among them, caerin 1.1, maximin 3, and dermaseptin S4 have been found to suppress HIV proliferation by direct inactivation.[17−19] However, no amphibian lectin has yet shown anti-HIV activity. Here, we identify and describe a novel lectin-like peptide, which we call fejerlectin, from the skin of Fejervarya limnocharis frogs. A series of structural analyses and pharmacological investigations demonstrate that fejerlectin is the smallest lectin-like peptide with potent agglutination and anti-HIV-1 activity identified to date.

Results

Identification and Characterization of Fejerlectin

Using polymerase chain reaction (PCR)-based cDNA cloning, we first obtained the complete nucleotide sequence encoding the fejerlectin precursor from a skin-derived F. limnocharis cDNA library. The nucleotide sequence has been deposited in the GenBank database under the accession code MW368972. As shown in Figure A, its precursor deduced from the 306 bp nucleotide sequence comprised 73 amino acid residues and contained the typical primary structure characteristic of amphibian defense peptides, with a signal peptide region, a N-terminal acidic spacer domain followed by a well-known KR protease cleavage site, and a mature peptide at the C-terminus.[20] Thus, the amino acid sequence of mature fejerlectin was predicted to be RLCYMVLPCP and contain one intramolecular disulfide bridge formed by two cysteines. The NCBI BLAST search did not find any peptide similar to the putative fejerlectin, suggesting that this peptide represented a new amphibian peptide family. Its theoretical isoelectric point and molecular weight were 8.01 and 1193.54 Da, respectively. Finally, the synthesized peptide was purified by high-performance liquid chromatography (HPLC) and confirmed by mass spectrometry, which was then used in subsequent experiments (Figure B,C).

Figure 1

Identification and characterization of fejerlectin. (A) cDNA and the deduced amino acid sequence of fejerlectin. The signal peptide is shaded in gray and is followed by an acidic spacer domain with KR residues at the end (in red bold). The stop codon is indicated with an asterisk (*), and the sequence of mature fejerlectin is boxed. (B) Purity of synthesized fejerlectin detected by HPLC. (C) Molecular weight of synthesized fejerlectin confirmed by mass spectrometry.

Hemagglutination (HA) Activity of Fejerlectin

The HA activity of fejerlectin is shown in Table . Fejerlectin could strongly agglutinate intact mice erythrocytes at a minimum concentration of 2.5 μM (8-fold dilution). The tested temperatures and pH did not affect its HA activity, indicating that fejerlectin was relatively stable under these conditions. Consistent with this, the HA activity of fejerlectin was also stable for 3 h in human plasma. Ethylenediaminetetraacetic acid (EDTA) treatment or addition of metal cations such as Ca2+ and Mg2+ had no effect on fejerlectin activity, suggesting that fejerlectin did not depend on metal cations to exert its lectin-like activity.

Table 1

HA Activity of Fejerlectin under Different Conditionsa

		HA units
test		1	2	4	8	16	32	64
	control	–	–	–	–	–	–	–
	fejerlectin	+	+	+	+	–	–	–
temperature (°C)	25	+	+	+	+	–	–	–
	37	+	+	+	+	–	–	–
	50	+	+	+	+	–	–	–
pH	6.5	+	+	+	+	–	–	–
	7.5	+	+	+	+	–	–	–
	8.5	+	+	+	–	–	–	–
serum stability (h)	1	+	+	+	+	–	–	–
	2	+	+	+	+	–	–	–
	3	+	+	+	+	–	–	–
metal ion	EDTA	+	+	+	+	–	–	–
	Ca2+	+	+	+	+	–	–	–
	Mg2+	+	+	+	+	–	–	–

+, hemagglutinating; −, not hemagglutinating.

Given the fejerlectin precursor’s structural similarity with amphibian defense peptides, its microbe-binding, microbe-killing, microbe-agglutinating, histamine-releasing, and mast cell degranulating activities, which are related to innate immunity, were investigated. Although fejerlectin did not show direct antibacterial activity (Figure S1), it could concentration-dependently bind and agglutinate bacteria (Figure B,C). In addition, fejerlectin at 125 μM promoted histamine release and degranulation of mast cells by 35.41 ± 4.4 and 26.93 ± 6.92%, respectively.

Figure 2

Binding reaction of fejerlectin with d-(+)-galacturonic acid. (A) Effects of d-(+)-galacturonic acid on the HA activity of fejerlectin. The second row shows the HA activity of fejerlectin at final concentrations between 20 and 0.3125 μM. The third row shows the HA activity of different concentrations of d-(+)-galacturonic acid (from left to right: 8, 4, 2, 1, 0.5, 0.25, and 0.125 mM). The fourth row shows the HA activity of 20 μM fejerlectin incubated with d-(+)-galacturonic acid at concentrations corresponding to the third row. (B) Flow cytometry of the binding reaction between fejerlectin and bacteria. Staphylococcus aureus and Escherichia coli were incubated with fluorescein isothiocyanate (FITC)-fejerlectin (3.75, 7.5, 15, and 30 μM) at 37 °C for 15 min before flow cytometry analysis. (C) Bacterial agglutination induced by fejerlectin. S. aureus and E. coli diluted to 2 × 108 cells/mL in Tris-buffered saline (TBS) were incubated with bovine serum albumin (BSA) (a, d), 5 μM fejerlectin (b, e), or 5 μM fejerlectin premixed with an equal volume of 4 mM d-(+)-galacturonic acid (c, f) for 1 h at room temperature and then stained with Gram dye. (D) Isothermal titration calorimetry (ITC) analysis of binding reaction of fejerlectin with d-(+)-galacturonic at 25 °C. The top panels displayed thermo changes of each injection at different time points, while the bottom panel presented the change of enthalpy as a function of ligand/target molar ratio. (E) Surface plasmon resonance imaging (SPRi) analysis of d-(+)-galacturonic acid binding to fejerlectin immobilized on a gold chip. Data were fit using a single-site binding model using the MicroCal Origin software package.

Carbohydrate-Binding Specificity of Fejerlectin

To investigate the carbohydrate-binding specificity of fejerlectin, a hemagglutination inhibition test was carried out. Of the 25 tested monomeric sugars, only d-(+)-galacturonic acid inhibited its HA and microbe-agglutinating activities (Figure A,C and Table S1), demonstrating that d-(+)-galacturonic acid might be the specific target of fejerlectin. Isothermal titration calorimetry (ITC) analysis further displayed that fejerlectin could bind d-(+)-galacturonic acid with the KD value of 4.39 × 10–6 M (Figure D and Table S2). In agreement, the binding reaction between fejerlectin and d-(+)-galacturonic acid was further demonstrated by surface plasmon resonance imaging (SPRi) experiment (Figure E). Taken together, all of those results further clarified that d-(+)-galacturonic acid was the specific target of fejerlectin.

Suppression of Infection by HIV-1 Infectious Clones

As shown in Figure A–C, like positive maraviroc (MVC), AMD3100, and AZT controls, fejerlectin possessed highly potent anti-HIV-1 activity against infection with HIV-1 infectious clones including HIV-1SF162 (R5), HIV-1NL4-3 (X4), and HIV-181A and NL4-3 (X4R5) with IC50 values of 2.17 ± 1.03, 1.43 ± 0.34, and 2.80 ± 0.37 μM, respectively (Figure A–C). Next, the potential cytotoxic effect of fejerlectin on TZM-bl cells was assessed to demonstrate that its anti-HIV-1 activity was not attributable to cytotoxicity. The 50% cytotoxicity concentrations (CC50) value of fejerlectin against TZM-bl cells was 182.49 ± 9.35 μM, which was much higher than the IC50 value of anti-HIV-1 activity (Figure D). These results indicate that fejerlectin has potent and broad-spectrum antiviral activity.

Figure 3

Anti-HIV-1 activity of fejerlectin and its cytotoxicity to host cells. The inhibitory activities of fejerlectin against HIV-1 infectious clones including HIV-1SF162 (A), HIV-1NL4-3 (B), and HIV-181A and NL4-3 (C). Maraviroc, AMD3100, and AZT were used as positive controls, respectively. (D) In vitro cytotoxicity of fejerlectin on TZM-bl cells. Experimental data are expressed as mean ± standard deviation (SD) (n = 3).

Effects on Early-Stage HIV-1 Infection

To understand on which stage of HIV-1 infection fejerlectin acts, a time-of-addition assay was performed. As shown in Figure , the three positive control drugs had strong inhibitory activity in early-stage HIV-1 infection. The anti-HIV activity of maraviroc and AMD3100 but not AZT significantly decreased 8 h after infection (Figure A,B). By contrast, AZT still had high antiviral activity when added 8 h after infection (Figure A–C). When 25 μM fejerlectin was added to the target cells before viral infection or 0.5–8 h post-infection, it also displayed obvious anti-HIV-1 activity. However, as the time of viral infection increased, its antiviral activity gradually reduced. In particular, when added 6 or 8 h after infection, the anti-HIV activity of fejerlectin against the three tested pseudotyped viruses decreased to less than 50% (Figure A–C). These results suggest that fejerlectin might exert its antiviral activities at an early stage of HIV-1 infection and be an HIV-1 entry inhibitor.

Figure 4

Time-of-addition assay with fejerlectin: (A) R5-monotropic HIV-1SF162, (B) X4-monotropic HIV-1NL4-3, and (C) X4R5 dual-tropic HIV-181A and NL4-3. Experimental data are expressed as mean ± SD (n = 3).

Effects on Single-Cycle Infection by HIV-1 Env-Pseudotyped Viruses

To further identify whether fejerlectin acts as an HIV entry/fusion inhibitor, a single-round entry assay was performed with maraviroc and AMD3100 as positive controls. Fejerlectin potently inhibited infection by all tested HIV-1 Env-pseudotyped viruses with EC50 values against HIV-1JR-FL and HIV-1HXB2 of 4.20 ± 0.96 and 9.63 ± 2.30 μM, respectively (Figure A,B). Moreover, VSV-G-pseudotyped virus expressing VSV-G Env as a negative control was used to examine the specificity of fejerlectin on HIV-1 Env. As shown in Figure C, fejerlectin had no inhibitory activity against VSV-G pseudovirus infection. These results suggest that fejerlectin might be an HIV-1 entry inhibitor that specifically targets HIV-1 Env membrane protein.

Figure 5

Effect of fejerlectin on the HIV-1 infection cycle. Inhibitory activities of fejerlectin against HIV-1JR-FL (A), HIV-1HXB2 (B), and VSV-G (C) pseudotyped viruses. Maraviroc, AMD3100, and T-20 were used as positive controls. (D) Effects of fejerlectin on the formation of syncytia between CHO-WT and MT-2 cells. ADS-J1 was chosen as a positive control. Experimental data are expressed as mean ± SD (n = 3).

Inhibition of HIV-1 Env-Mediated Cell–Cell Fusion

Env-mediated membrane fusion is pivotal for HIV-1’s early entry into a target cell. The above results suggest that the anti-HIV-1 activity of fejerlectin could be attributable to targeting HIV-1 Env. Therefore, we further examined the inhibitory activity of fejerlectin on HIV-1 Env-mediated cell–cell fusion with CHO-WT cells as the effector cells and MT-2 cells as the target cells.[21] As shown in Figure D, fejerlectin dose-dependently inhibited cell–cell fusion with an IC50 value of 6.57 ± 0.25 μM, which was comparable to that of the ADS-J1-positive control. This result further suggests that fejerlectin might inhibit HIV-1 entry by blocking HIV-1 Env-mediated membrane fusion.

Inhibition of HIV-1 Gp41 6-HB Formation

Gp41 exerts a crucial role during fusion of the viral and cellular membranes.[22] The HIV-1 gp41 6-HB is composed of N- and C-terminal heptad repeat sequences (NHR and CHR), a key structure enhancing HIV-1 fusion and entry into target cells. Gp41 core formation in vitro was mimicked with the N- and C-peptides based on an effective model system established in our laboratory.[23,24] Both fejerlectin and ADS-J1 significantly inhibited gp41 6-HB formation in a dose-dependent manner with IC50 values of 2.61 ± 0.17 and 1.53 ± 0.48 μM, respectively (Figure A). In agreement, as shown by circular dichroism (CD) spectroscopy, the conformational formation of α-helicity between the N36 and C34 peptide was inhibited by incubating various concentrations of fejerlectin with N36 before addition of C34 peptides (Figure B). Therefore, fejerlectin inhibited the formation of α-helical complexes and the 6-HB structure by blocking the interaction between the viral gp41 NHR and CHR regions. To further ensure that fejerlectin bound N36, the kinetic reaction between fejerlectin and N36 was analyzed by SPRi. Fejerlectin strongly bound to N36 peptide with a KD of 4.66 × 10–4 M (Figure C). Taken together, these results suggest that fejerlectin blocks the formation of gp41 6-HB and inhibits entry of HIV-1 into target cells.

Figure 6

Effect of fejerlectin on the formation of 6-HB. The inhibitory effect of fejerlectin on the formation of 6-HB was analyzed by enzyme-linked immunosorbent assay (ELISA) (A) and CD spectra (B). ADS-J1 was chosen as a positive control. Experimental data are expressed as mean ± SD (n = 3). (C) Binding affinity between fejerlectin and N36 peptide derived from HIV-1JR-FL gp41 analyzed by SPRi.

Discussion

Amphibians produce a variety of defense molecules, such as antimicrobial peptides and lectins, in their skin to protect them from invading microbes and predators.[16] Although a common amphibian species in Asia, only a few studies have examined the activity of skin secretions from F. limnocharis, and reports on its defense peptides are limited.[25,26] Here, for the first time, we identify and describe a novel lectin-like peptide called fejerlectin from F. limnocharis skin. Although its precursor shares some structural similarities to other amphibian defense peptides and its mature peptide contains a disulfide-bridged cyclic region comprising seven amino acid residues present in some known antimicrobial peptides (Figure A), fejerlectin has no antimicrobial activity (Figure S). However, the 10 amino acid fejerlectin peptide shows significant hemagglutination and d-(+)-galacturonic acid-binding activities. Moreover, fejerlectin has significant anti-HIV-1 activities by blocking the formation of gp41 6-HB and inhibiting HIV-1 entry into target cells. Therefore, fejerlectin is the smallest lectin-like peptide reported to date and represents a new family of lectins.

Lectins are a class of non-immune-origin proteins that can specifically bind to carbohydrates and that are involved in development, infectious diseases, and immunological functions, making them suitable as drugs and drug carriers.[27,28] However, most lectins are large and cannot be used as drugs due to their immunogenicity and toxicity. Smaller peptides or even organic molecules that can mimic the function of lectins are ideal drug candidates.[29] However, only a limited number of small peptide lectins have shown anti-HIV activity. Moreover, the reported natural peptide lectin with anti-HIV activity contains more than 10 residues and more complex in structure.[7] For example, retrocyclin-1 is a circular octadecapeptide with an internal tridisulfide ladder.[13] Thus, it is overburdened and expensive for current production systems if they are used as anti-HIV drugs. As the smallest lectin with a single disulfide bridge in its structure, fejerlectin should be stable and easy to synthesize. Consistent with this, fejerlectin is easily synthesized (Figure B,C) and a small range of temperatures, pH changes, and serum incubation conditions do not alter its HA activity (Figure and Table ). Furthermore, fejerlectin has low toxicity and can suppress early viral entry into host cells at the sites of infection, making it desirable for development into a microbicide like most naturally occurring anti-HIV peptides (Figure ). Mutations in the envelope proteins of HIV-1 can easily result in resistance to many drugs targeting gp120 but not the molecules binding to carbohydrates.[30] Therefore, fejerlectin, with its lectin-like activity, can easily retain broad-spectrum anti-HIV activity. In addition, several protein lectins have reduced HIV transmission in organotypic models, suggesting that naturally occurring lectins are promising anti-HIV drugs.[31,32] It is worth noting that such a small peptide can be so powerful to agglutinate bacterial cells and erythrocytes. The N value measured in ITC is 0.415, indicating that one d-(+)-galacturonic acid may bind to multiple fejerlectin molecules (Table S2). Furtherly, the size distribution of fejerlectin is obviously different in number and intensity (Figure S2), indicating that fejerlectin agglomerates in the solution and may exist in the form of a larger complex.[33] Considering that more than one target sugar molecule is on cell surface, fejerlectin serves as a bridge and its cross-link in solution plays an important role during cell agglutination. Overall, the small size, stable properties, low toxicity, and significant antiviral activities make fejerlectin an intriguing prototype for the development of an ideal anti-HIV drug.

HIV-1 entry into host cells mainly depends on Env-mediated membrane fusion. Consequently, HIV-1 Env, with its large number of glycosyl sites, is a target for anti-HIV-1 lectins.[12,34] As a lectin-like peptide, fejerlectin inhibits HIV entry by suppressing Env-mediated membrane fusion during HIV infection, and indeed our compelling results support this hypothesis (Figure ). By contrast, almost all anti-HIV peptides from amphibian skin exert their functions through direct inactivation involving their oligomerization in bilayers and subsequent disruption of membrane integrity, as demonstrated by caerins and dermaseptins.[18,19,35,36] Thus, fejerlectin is the first amphibian-derived peptide to directly interfere with HIV entry. Surprisingly, fejerlectin can bind directly nonglycosylated N36 peptide and inhibit the formation of 6-HB (Figure ). Furthermore, to the best of our knowledge, the presence of d-(+)-galacturonic acid in Env has yet to be reported. Therefore, the anti-HIV activity of fejerlectin seems to be unrelated to its carbohydrate-binding activity. In recent years, distinct classes of HIV-1 entry inhibitors targeting 6-HB of gp41 have been discovered. For example, our previously studies confirmed that a glycosylated dihydrochalcone derived from natural products (trilobatin) and a N-substituted pyrrole derivative 12 m (NSPD-12 m) inhibited HIV-1 entry/fusion by binding the gp41 NHR region and blocking the formation of fusion-active 6-HB.[24,37] In addition, there are also a large number of literature reports on peptide-based HIV entry/fusion inhibitor targeting gp41. The first peptide-based HIV entry inhibitor, enfuvirtide (T20), was approved by Food and Drug Administration (FDA) to treat HIV/AIDS patients in 2003.[38] The second-generation CHR-derived peptides in a hope to replace T20 have potent anti-HIV-1 activity by inhibiting 6-HB formation, such as sifuvirtide and albuvirtide.[39,40] The third-generation CHR peptides are CHR-derived peptides containing MT-hook or IDL anchor, and lipopeptides with 50-fold more potent anti-HIV-1 activity than T20 including CP32, HP23, and CP-IDL.[41,42] However, our data support it is possible that fejerlectin can bind both carbohydrates and the gp41 NHR.

Previous studies have shown that mannose, the target of anti-HIV lectins, is mainly present in gp120.[43−45] However, fetuin-specific retrocyclin-1 can protect cells from HIV-1 infection by binding with high affinity to gp120, CD4, and galactosylceramide rather than nonglycosylated gp120.[13] Furthermore, gp41 contains a large amount of complex-type glycans other than mannose and can be bound by the β-galactose-specific lectin CVL, suggesting that other monosaccharides or glycoproteins in HIV-1 Env may also serve as targets for molecules inhibiting HIV-1 entry.[9,46,47] Considering that we only investigated a limited number of monosaccharides and fejerlectin inhibits Env-mediated fusion, we cannot exclude that other sugar molecules or glycoproteins in Env are fejerlectin targets and related to its anti-HIV activity.

In summary, this is the first report identifying a lectin-like peptide from the skin of F. limnocharis. The primary sequence of fejerlectin contains 10 amino acid residues including a disulfide-bridged heptapeptide ring. Fejerlectin has significant hemagglutination and d-(+)-galacturonic acid-binding activities. Thus, it is the smallest lectin-like peptide reported to date. Furthermore, fejerlectin can prevent the formation of 6-HB by binding to gp41 NHR, thus inhibiting Env-mediated fusion and HIV-1 entry. These characteristics make fejerlectin a promising new template for future anti-HIV drug development.

Materials and Methods

Animals and Ethics Statement

Adult F. limnocharis frogs of both genders weighing 3–6 g (n = 4), which are not an endangered or protected species, were captured in the countryside of Guangzhou, Guangdong Province, China. The frogs were humanely euthanized by CO2 inhalation, washed with deionized water, and their skin was separated immediately for RNA extraction. All protocols and procedures involving live animals were approved by the Animal Care and Use Ethics of Southern Medical University and carried out in full compliance with the guidelines approved for Animal Care and Use at Southern Medical University.

Reagents

HEK-293T, TZM-bl, CHO-WT, and MT-2 cells were obtained from the American Type Culture Collection. pNL4-3E-R-Luc plasmid, HIV-1 and vesicular stomatitis virus-G (VSV-G) Env-encoding plasmids, maraviroc (MVC, a CCR5 antagonist), AMD3100 (a CXCR4 inhibitor), and zidovudine (AZT, a nucleoside reverse transcriptase inhibitor) were purchased from the National Institutes of Health AIDS Research and Reference Reagent Program. Plasmids encoding HIV-1JR-FL (R5 strain), HIV-1HXB2 (X4 strain), HIV-1NL4-3 (CXCR4-tropic), HIV-1SF162 (CCR5-tropic), and dual-tropic HIV-181A and NL4-3 infectious clones were kind gifts from Jan Munch of Ulm University. All virus stocks were produced by transfecting HEK-293T cells. The well-characterized peptides N36 and C34 derived from the N- and C-terminal core structure of HIV-1 gp41 were synthesized by Scilight Biotechnology LLC (Beijing, China).[23,48] ADS-J1, a small-molecule HIV-1 entry inhibitor that blocks gp41 six-helix bundle (6-HB) formation, was purchased from ComGenex. Mouse monoclonal antibodies targeting NC-1 and T-20 specific for the gp41 6-HB were prepared using a previously described method.[49−51]

Molecular Cloning and Characterization of cDNA Encoding Fejerlectin

Total RNA from F. limnocharis skin was extracted using TRIzol reagent (Life Technologies, Inc.), purified by oligo (dT) cellulose chromatography (Life Technologies, Inc.), quantified with a Merinton SMA1000 (Merinton), and used for cDNA synthesis with a SMART cDNA library construction kit (Takara Biotechnology) according to the manufacturer’s instructions. cDNA encoding fejerlectin was obtained by PCR amplifications with synthesized skin total cDNA as template and HG (5′-AGATGTT(G/C)ACC(T/A)TGAAGAAATC-3′) in the sense direction plus 3′ PCR primer CDS III (5′-ATTCTAGAGGCCGAGGCGGCCGACATG-3′) in the antisense direction as primers.[52] PCR amplifications were carried out in a reaction mixture containing Gene Taq polymerase (TianGen) in a thermal cycler. The PCR program was as follows: 4 min at 95 °C; 30 cycles of 20 s at 95 °C, 30 s at 48 °C, and 30 s at 72 °C, and finally 10 min at 72 °C for extension. The PCR product (∼300 bp) was purified by gel electrophoresis and cloned into the pMD18-T vector (Takara Biotechnology) for DNA sequencing. The physical and chemical parameters of fejerlectin and its precursor were predicted and analyzed using the ExPASy Bioinformatics Resource Portal (http://www.expasy.org/tools/) and the blastp suite of the Basic Local Alignment Search Tool (BLAST) at the National Center for Biotechnology Information (NCBI) (https://blast.ncbi.nlm.nih.gov/blast.cgi).

Peptide Synthesis

Fejerlectin and FITC-labeled fejerlectin (FITC conjugated to the N terminus of fejerlectin) were synthesized by GL Biochem Ltd. The crude synthetic peptide was purified with an Inertsil ODS-SP (C-18) RP-HPLC column (Shimadzu) to >95% purity. The high-purity peptide was pooled, lyophilized, and further characterized by matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectrometry.

Hemagglutination Activity Test

Mice erythrocytes were washed with Alsever’s solution and diluted into a 2% suspension (v/v) with phosphate-buffered saline (PBS, pH 7.5). HA activity tests were carried out in V-bottom microtiter plates by the twofold dilution method. After erythrocytes in the blank had fully subsided, the results were recorded with PBS as a negative control. The HA titer, defined as the reciprocal of the lectin displaying HA activity at the lowest concentration, was considered as one HA unit.[53] To determine the effects of temperature, pH, serum, metal ions, and carbohydrates on its HA activity, fejerlectin was incubated at three different temperatures (25, 37, and 50 °C); in selected buffers (100 mM sodium citrate, pH 6.5; 100 mM Tris–HCl, pH 7.5; 100 mM glycine, pH 8.5); with PBS with 10 mM EDTA, 10 mM CaCl2, or 10 mM MgCl2; in human serum at 37 °C for 0–4 h; or with PBS containing 25 different sugars including d-(+)-galacturonic acid for 1 h before HA titers were measured. All experiments were repeated at least three times.

Bacterial Agglutination Assay

The agglutination tests against S. aureus and E. coli were performed to further verify the lectin-like activity of fejerlectin. S. aureus and E. coli were diluted to a density of 2 × 108 cells/mL and incubated at room temperature for 1 h with BSA, 5 μM fejerlectin, or 5 μM fejerlectin premixed with 4 mM d-(+)-galacturonic acid for 30 min, respectively. Then, the mixtures were individually stained with Gram dye and observed under an inverted microscope.

Isothermal Titration Calorimetry

The interaction between limnolectin and d-(+)-galacturonic acid was measured with a VP-ITC microcalorimeter (Malvern, U.K.) as previously reported by us.[52] In short, limnolectin and d-(+)-galacturonic acid were dissolved in 50 mM PBS, pH 7.2, and degassed under vacuum before use. d-(+)-Galacturonic acid (10 mM) was added in 1.0 μL aliquots every 2 min intervals into 50 μM limnolectin in the sample cell under constant stirring at 25 °C for 21 injections. PBS solution was used as a blank control to subtract the dilution heat. High feedback mode of the instrument was selected to analyze the data. The enthalpy change (ΔH) and the equilibrium dissociation constant (KD) were calculated by fitting to a single-site binding model with the MicroCal Origin software. The basic thermodynamic equations were used to calculate the entropy change (ΔS) and Gibb’s free-energy change (ΔG). The experiment was repeated at least three times.

Surface Plasmon Resonance Imaging Assay

The affinity between fejerlectin and d-(+)-galacturonic acid was investigated with the PlexArray HT system. Briefly, fejerlectin (2 mM) was fixed on the chip surface and then incubated at 4 °C overnight in a humid box. Before SPRi, the chip was washed and blocked with 5% (m/v) nonfat milk overnight. After the baseline was stabilized with PBS, different concentrations of d-(+)-galacturonic acid (10, 20, and 40 mM) or N36 peptide (125, 250, and 500 μM) were flowed over the chip at 2 μL/s. H3PO4 (0.5% (v/v)) in deionized water was used for chip surface regeneration. Standard deviations were plotted using the average results of 10 different points, and the average kinetics of the complexes were obtained from the 1:1 Langmuir kinetic model of three different concentrations. The KD value was calculated according to the kinetic constants derived from curve-fitting association and dissociation rates to the real-time binding and washing data. All data analysis was performed with ORIGINLab software. The experiment was repeated at least three times.

Anti-HIV-1 Infection Assay

The anti-HIV-1 infection activity of fejerlectin was measured with three different HIV-1 infectious clones (CXCR4-tropic HIV-1NL4-3, CCR5-tropic HIV-1SF162, and dual-tropic HIV-181A and NL4-3) and HIV-1NL4-3-Luc virions. First, TZM-bl cells (1 × 105/mL) were seeded into 96-well plates and incubated at 37 °C overnight. Then, 100-fold 50% tissue culture infective dose (TCID50) HIV-1 or VSV-G viruses were incubated with fejerlectin at various concentrations at 37 °C for 30 min. Then, the mixture was added to cells and further incubated for 48 h. Finally, a Dual-Luciferase Reporter Assay Kit was used to determine the luciferase activity of lysed TZM-bl cells. Maraviroc, AMD3100, and AZT were used as positive controls, respectively. For the single-round HIV-1 infection assay, TZM-bl cells were challenged with HIV-1 Env-pseudotyped HIV-1JR-FL (R5 strain) and HIV-1HXB2 (X4 strain) viruses, which were preincubated with fejerlectin (0.5–50 μM) at 37 °C for 30 min. Fresh Dulbecco’s modified Eagle’s medium (DMEM) was used to replace the culture supernatant at 24 h after viral infection, and the cells were collected 72 h post-infection for luciferase activity measurement.

Cytotoxicity In Vitro

The in vitro TZM-bl cell cytotoxicity of fejerlectin was measured by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. Briefly, TZM-bl cells (2 × 104/well) were inoculated into 96-well culture plates containing DMEM with 1% penicillin/streptomycin and 10% fetal bovine serum. After 24 h, graded concentrations of fejerlectin from 1 to 1000 μM were added to the medium prior to another 48 h of incubation. The medium was removed, and 100 μL of MTT solution (0.5 mg/mL) was applied to each well. After 4 h, the MTT solution was replaced with 150 μL of dimethyl sulfoxide (DMSO) to dissolve the formazan crystals. Absorbance at 570 nm was determined with an ELISA reader. The 50% cytotoxicity concentrations (CC50) were computed with CALCUSYN software.

Time-of-Addition Assay

To investigate the in vitro antiviral activity of fejerlectin at different time points after virus infection, the time-of-addition assay was carried out as previously reported.[54] In short, 50 μL of HIV-1NL4-3, HIV-1SF162, and HIV-181A and NL4-3 at 100 TCID50 were used to infect 100 μL of 5 × 105/mL TZM-bl cells for 0, 0.5, 1, 2, 4, 6, 8, and 24 h at 37 °C before adding 50 μL of fejerlectin (25 μM), AZT (45 nM), maraviroc (24 nM), or AMD3100 (50 nM). The culture supernatants were replaced with fresh medium 24 h post-infection, and the culture supernatants were collected for measuring p24 antigen levels, as described in the above anti-HIV-1 infection assay after 48 h of infection.

HIV-1 Env-Mediated Cell–Cell Fusion Assay

The syncytium formation assay was performed using our previous method to directly evaluate the inhibition of fusion between MT-2 cells (expressing CD4 and CXCR4) and CHO-WT cells (expressing HIV-1 Env).[21] Briefly, 2 × 105/mL CHO-WT cells were incubated with graded concentrations of fejerlectin for 30 min and then incubated with 4 × 105/mL MT-2 cells for 48 h at 37 °C. ADS-J1 was used as a positive control. The number of positive syncytia formed by MT-2 and CHO-WT cells in three fields per well was recorded using an inverted microscope. The percentage inhibition of cell fusion was computed according to the following formula: %inhibition = (1 – syncytia number of inhibitor/syncytia number of positive control) × 100. The IC50 values were counted with CALCUSYN software.

Circular Dichroism Spectroscopy Assay

N36 at 10 μM was incubated with fejerlectin at a final concentration of 5, 6.25, 12.5, 25, or 50 μM in PBS at 37 °C for 30 min before incubation with 10 μM C34 for another 30 min. The CD spectra of all samples from 190 to 260 nm were measured with a Chirascan CD spectrometer (Applied Photophysics Ltd., U.K.) using a 5 nm bandwidth, 0.1 cm path length, 0.1 nm resolution, 4 s response time, and 50 nm/min scanning speed. The data were corrected by subtraction of the blank corresponding to the solvent and are presented as the mean residue ellipticity (θ) in deg cm2/dmol.[55]

ELISA for Inhibiting Gp41 6-HB Formation

To determine whether fejerlectin interfered with gp41 6-HB formation, a sandwich ELISA was established as described previously.[56] First, 2 μM N36 peptide was incubated with a graded concentration of fejerlectin (5–50 μM) at 37 °C for 30 min before further incubation with 2 μM C34 at 37 °C for another 30 min. Then, the mixture was added to a 96-well polystyrene plate precoated with 2 μg/mL rabbit anti-6-HB IgG. Then, 1 μg/mL NC-1 monoclonal antibody, biotin-labeled anti-mouse IgG (Sigma-Aldrich; 1:5000 diluted in 0.01 M PBS with 2% dry nonfat milk), streptavidin-labeled horseradish peroxidase (HRP) (Zymed, South San Francisco, CA; 1:10 000 diluted in PBS with 10% goat serum), 3,3′,5,5′-tetramethylbenzidine, and 1 M H2SO4 were sequentially added. Finally, the absorbance value at 450 nm was measured with an ELISA reader.