Two Hypothetical Proteins from Bacillus thuringiensis Genomes were Parasporinlike Proteins

 Research Article

 Two Hypothetical Proteins from Bacillus thuringiensis Genomes were Parasporinlike Proteins

Corresponding author:Dr. Lin Yi, Department of Bioengineering & Biotechnology, College of Chemical Engineering, Huaqiao University, Jimeidadao 668#, Jimei, Xiamen 361021, Fujian, China, Tel: 086-592-6166131; 


Parasporins are a group of anti-cancer proteins from Bacillus thuringiensis that preferentially kill cancer cells leaving the normal cells unaffected. Two hypothetical proteins were found from B. thuringiensis genomes to be homologous to Parasporin-2. Their structures were both β-sheet-dominant one with 4 α-helix occurred only in domain I, which were also seen in Parasporin-2. An aromatic cluster in domain I of the two proteins might serve as the binding site for target cells’ receptors as reported for Parasporin-2. A Ser/Thr track that flanks domain I and domain II might contribute to theoligomerization of the protein and further enables the proteins to move laterally across the target cell membrane was also identified in the two hypothetical proteins. They also shared the same or similar key residues with Parasporin-2.

Keywords: Parasporin; Bt Genome; Hypothetical Proteins; Aromatic Cluster; Ser/Thr Track; Key Residues


Bacillus thuringiensis, a gram-positive and endospore-forming bacterium, produce large crystalline inclusion bodies during sporulation [1,2]. Often, the crystal proteins in the inclusion bodies are highly toxic to agriculturally and medicallyimportant insects, including Lipdoptera, Diptera, Coleoptera and Hemiptera [1]. However, previous studies reported someBt strains contain proteins that are both non-insecticidal and non-hemolytic, but can preferentially kill human cancers.These proteins are referred to as parasporins (PS) [2]. Parasporins were classified into six groups based on their sequence identities according to the Parasporin Classification and nomenclature Committee ( They were all isolated from from Bt strains and showed different anticancer activities. Parasporin-2 is the only one with an identified X-ray structure among all parasporins (PDB ID: 2ZTB). A new way of mining potent parasporins from the currentdozens of Bt genomes was presented in this paper. We found two hypothetical proteins from Bt genomes with high structural similarities to that of parasporin-2.

Materials and Methods
Genome Mining for Potent Parasporin-like Proteins
The amino acid sequence of the active Parasporin-2 (UniProt accession number: B5BUV6) was used to search for similarsequences by blasting against the genome of Bacillus thuringiensis using Protein Blast. Proteins with an identity less than45% and are not annotated with any specific functions were retrieved. Two hypothetical proteins showing high structuralsimilarities with Parasporin-2 (PDB ID:2ZTB) were chosen for present study.

Protein Structure Analysis
The retrieved sequences were submitted to the Phyre2 server for structure modeling. Ramachandran plots were applied for evaluating protein structures and were generated by RAMPAGE server. Protein structure similarity search was done by submitting the hypothetical protein structures to the Dali server. For comparison, results with a Z-score greater than 10 and less than 10 were chosen. Protein structure profiles were retrieved from the Protien Data Bank (PDB). Visualization and comparison of protein structures were done by PYMOL. Multiple sequence alignment was done by DNAMAN.

Results and Discussion

Two Hypothetical Proteins Homologous to Parasporin-2 were Found from Bt Genomes
In order to find out potent anti-cancer proteins from the genomes of Bacillus thuringiensis, amino acid sequence of the active Parasporin-2 was used for blasting against the Bt genomes. Two hypothetical proteins were found from the Bt genomes, referred to as Scan-1 (accession number: WP_000586619.1 ) and Scan-2 (accession number: WP_000586616.1), showing 32% and 33% homologies to Parasporin-2, respectively (Figure 1A).

Figure 1. Multiple Sequence Alignment (MSA)Results.
(A) Partial MSA results of Parasporin-2 and two hypothetical proteins: Scan-1 and Scan-2 (Black: highly conserved, Blue: moderately conserved).
(B) Partial MSA results of Parasporin2, Scan-1, Scan-2, epsilon toxin and 26kDa non-toxic protein (Blue: moderately conserved, Magenta: conserved).

bio onfo fig3.1

Scan-1 contains 253 residues with a putative molecular mass of 28.2 kDa, Scan-2 is also a 253-residue-protein with a putative molecular mass of 28.3 kDa, which showed similar properties with the active form of Parasporin-2 (30kDa).

The two Hypothetical Proteins Showed Similar 3-D Structure to Parasporin-2
Protein sequences of the two hypothetical proteins were submitted to the Phyre2 server for structural modeling. Results were then evaluated using Ramachandran plot. For Scan-1, 235 residues (97.1%) were in favored region, with 5 and 2 residues in the allowed and outlier region, respectively (data not shown).For scan-2, 233 residues (96.3%) were in favored region, with 5 and 4 residues in the allowed and outlier region, respectively (data not shown). The results showed that the models were reliable.Results showed that the structure of Scan-1 and Scan-2 were both β-sheet-dominant one with 4 short alpha helices forming the “head” of the protein (Figure 2bc). The proteins showed a similar three-domain structure with Parasporin-2 (Figure. 2a) [3].
Figure 2. Structures of 5 Toxins. Structures of 5 toxins. (a) Parasporin-2 (PDB ID: 2ZTB), (b) Scan-1, (c) Scan-2, (d) Epsilon toxin (PDB ID: 1UYJ), (e) 26NPT (PDB ID: 2D42).

 bio onfo fig3.1

Three of the alpha helices were located in domain I with one located at the interface of domain I and domain II. The secondary structure between Scan-1 and Scan-2 had minor difference at position 68-71 where a 4-aminoacid-β-sheet of Scan-1 is longer than the 3-aminoacid-β-sheet of Scan-2 (Figure 3bc).

The secondary structures of the two hypothetical proteins highly resemble that of Parasporin-2, major difference occurred at one β-sheet of Parasporin-2 (Figure 3a, Q119-T139). Scan1 and Scan-2 held two loop regions at the same position (Figure 3bc) flanking N71-N75 and K77-V80 for Scan-1, R70-P75 and I77-I80 for Scan-2, however, these regions contains no key residues predicted previously by our lab [4].

Figure 3. Major secondary structure differences between three proteins. (a) Parasporin-2, (b) Scan-1 and (c) Scan-2.

bio onfo fig3.1

The two Hypothetical Proteins Share Similar Functional Blocks with Parasporin-2.
In order to analyze the function of the hypothetical proteins, an InterProScan project was carried out. Results showed that both Scan-1 and Scan-2 were aerolysin-like toxins, with an aerolysin β-complex domain (data not shown). For further proof, a Dali search was carried out, using the hypothetical protein structure profiles to search for similar structure elements. For comparison, we chose both results with a Z-score greater and less than 10 for further analysis. Parasporin-2(PDB ID: 2ZTB) showed highest Z-score with two hypothetical toxins (31.2 for Scan-1 and Scan-2), two other toxins, Epsilon toxin (PDB ID: 1UYJ) showed a Z-score of 9.9 and 7.2 for Scan-1 and Scan-2 respectively, 26kDa non-toxic protein (26NPT, PDB ID: 2D42) showed a Z-score of 8.5 and 9.1 for Scan-1 and Scan-2, respectively. All 5 toxins showed a β-sheet dominant structure (Figure 2). Domain I consists of several short antiparallel β-sheet and 3-4 ɑ-helices, which was previously reported to be receptor binding domain, differences in this domain determine the target specificity of these toxins [3,5]. The conservation of this structure was in consistency with the conserved regions revealed by multiple sequence alignment (Figure 1B upper panel). Among the five toxins, Parasporin-2, Scan-1 and Scan-2 showed similar folding patterns in this domain with only minor differences between length of alpha helices located at the interface of domain I and domain II (Figure 2abc). Notably, we found a highly conserved block within the first 80 amino acid of the three proteins (Figure 1A upper panel, highlighted in black). These identical sequences between the three proteins revealed a same function of binding to the same target cell receptor patterns. Domain II and domain III were both consist of long antiparallel β-sheets that form a β-barrel which is a typical structure element in aerolysin-like toxins [3]. Multiple sequence alignment also showed that several conserved blocks occurred in these two domains (Figure 1B lower panel). The β-barrel was believed to be important function unit for carrying out their toxicity [3,6,7]

The Two Hypothetical Proteins Exhibited Similar Aromatic Cluster to Parasporin-2
Recent studies reported that the aromatic residues clustered in domain I of Parasporin-2 may serve as binding site for membrane receptors, which might be sugar chains of membrane proteins or lipid concentrated in the lipid rafts of target cells, as proposed for epsilon toxins [3]. In this study, we found two major aromatic clusters among the five proteins, one located in domain I (Figure 4B), another located in the β-barrel of domain II (Figure 4C).

Figure 4. Aromatic Cluster of 5 Toxins. (A) Aromatic clusters showed on surface of Parasporin-2, Scan-1, Scan-2, epsilon toxin, 26NPT (a-e, aquamarine), yellow patches indicate clusters occurred only on parasporin-2. (B) Aromatic residues in domain I of Parasporin-2, Scan-1 and Scan-2 (a-c, aquamarine), residues occurred only on Parasporin-2 were shown in yellow..

 bio onfo fig3.1

All five proteins showed a large continuous patch of aromatic residues in domain I. Only the 26kDa non-toxic protein held the cluster at the back of the molecule (Figure 4A e). Parasporin-2, Scan-1 and Scan-2 shared very similar distribution patterns of aromatic residues except for two patches (Figure 4A a, yellow). We further analyzed the aromatic residues located inside domain I of the three proteins. Most of the aromatic residues share similar folding patterns, which are in consistency with multiple sequence alignment results (Figure 1A upper panel). However, Scan-1 and Scan-2 lack some of the residues compared with Parasporin-2 (Figure 4B a, yellow), among them, Tyr57 (Figure 5B a, labeled red) was previously reported by our lab to be related to the activity of Parasporin-2 against liver cancer [4]. Compared with Parasporin-2, Scan-1 and Scan-2 both have one Phe (F220) located at the interface of domain I and domain II (Figure 4 B a). Compared with Parasporin-2, Scan-1 and Scan-2 both lack one Tyr at position 35 (Figure 4B abc). All these differences can be helpful for further explanation on the different activities of the three proteins.

By comparing the aromatic cluster in domain II, we found that most of the aromatic residues of Parasporin-2 and epsilon toxin were protruding toward the hydrophobic core of the β-barrel (Figure 4C), Scan-1 and Scan-2 showed the same distribution pattern. However, most aromatic residues in this domain of the 26kDa non-toxic protein were protruding outward except for a Phe (F164) and a Tyr (Y82) that fold toward the hydrophobic core (Figure 4C e), these hydrophobic residues may hinder the stability and toxicity of the 26kDa non-toxic protein. Parasporin-2 and the two hypothetical proteins had five residues (Figure 4C abc, yellow) that folded in highly consistent manner. The hydrophobic interaction between these aromatic residues that protruding inward might enhance the stability of the β barrel which is believed to be important elements involved in the oligomerization and pore formation [3,7].

The two Hypothetical Proteins have the Track of Ser/Thr Similar to Parasporin-2
The track of Ser/Thr was suggested to be involved in oligomerization of β-pore-forming toxins, it was also proposed by Toshihiko et al. that the track of Ser/Thr also enables the protein to move freely on the membrane of target cells [3]. Thereby, we analyzed the distribution pattern of Ser and Thr of the five proteins (Figure 5).

Parasporin-2 and epsilon toxin showed a long continuous Ser/Thr track that flanks domain II (Figure 5 a and d) while the 26kDa non-toxic protein showed a scattered distribution of Ser and Thr residues (Figure 5), this indicated a non-functioning structure of the protein. Scan-1 and Scan-2 showed a relatively continuous patch of Ser and Thr (Figure 5 b and c) compared with the 26kDa non-toxic protein, however, the patch covers a smaller surface than that of Parasporin-2 and epsilon toxin..

Figure 5. Tracks of Ser/Thr (Blue) of 5 toxins. (a) Paraporin-2 (PDB ID: 2ZTB), (b) Scan-1, (c) Scan-2, (d) Epsilon toxin (PDB ID: 1UYJ), (e) 26NPT (PDB ID: 2D42).

 bio onfo fig3.1

The two Hypothetical Proteins have the Same or Similar Key Residues of Parasporin-2 Related to the ActivitiesAgainst Liver Cancer.

We previously analyzed the key amino acid of parasporin-2 using molecular docking and reported the residues that affecting its activity against liver cancer cell (Table. 1) [4]. In the present study, the amino acid residues of two hypothetical proteins located at corresponding position to that of parasporin-2 were analyzed.

Table 1. Key functioning amino acid of Parasporin-2 and their counterparts of two hypothetical proteins.

 bio onfo fig3.1

Three residues of Scan-1 and Scan-2 were exactly the same as its counterpart of Parasporin-2 (Asn, Phe and Pro for Scan-1), which indicated the same function might be carried out by these residues. Others, different though, shares similar chemical properties: Thr-Ile, Ile-Phe, Pro-Val, Leu-Val are pairs that both have a hydrophobic side chain and a hydrophilic side chain for Thr-His.

In summary, the two hypothetical proteins found from B. thuringiensis genomes were Parasporin-like proteins based on structural analysis. According to the nomenclature, the current six Parasporins shared less than 45% homologies in amino acid sequence. The two proteins presented in this paper may be considered as the seventh or eighth Parasporin if they demonstrated anticancer activities…


1.Okassov A, Nersesyan A, Kitada S, Ilin A. Parasporins as new natural anticancer agents: a review. Journal of Buon. 2015, 20(1): 5-16.

2.Mizuki E, Park YS, Saitoh H, Yamashita S, Akao T et al. Parasporin, a human leukemic cell-recognizing parasporal protein of Bacillus thuringiensis. Clinical and Vaccine Immunology. 2000, 7(4): 625-634.

3.Akiba T, Abe Y, Kitada S, Kusaka Y, Ito A et al. Crystal structure of the parasporin-2 Bacillus thruingiensis toxin that recognizes cancer cell. J Mol Biol. 2009, 386(1): 121-133.

4.Liao LM, Lin SF, Tian L, Chen AM, Lin Y. Key aromatic amino acids of anti-hepatoma activity on parasorin-2. Chinese Journal of Biotechnology. 2013, 29(6): 823-835.

5.Kitada S, Abe Y, Maeda T, Shimada H. Parasporin-2 requires GPI-anchored proteins for the efficient cytocidal action to human hepatoma cells. Toxicology. 2009, 264(1-2): 80-88.

6.Mahalakshmi A, Shenbagarathai R. In silico structural analysis of parasporin 2 protein sequences of non- toxic Bacillus thuringiensis. Journal of Biomedical Science and Engineering. 2010, 3(4): 415-421.

7.Xu CC, Chinte U, Chen LC, Yao QQ, Meng Y et al. Crystal structure of Cry51Aa1: A potential novel insecticidal aerolysin-type β-pore-forming toxin from Bacillus thuringiensis. Biochemical and Biophysical Research Communications. 2015, 462(3): 184-189.

Be the first to comment on " Two Hypothetical Proteins from Bacillus thuringiensis Genomes were Parasporinlike Proteins"

Leave a comment

Your email address will not be published.