Skip to main page content

• HOME
• Current Issue
• Archive
• Alerts
• About ASM
• Contact us
• Tech Support
• Journals.ASM.org

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Dennis, D. Articles by Augustine, B. Search for Related Content PubMed PubMed Citation Articles by Dennis, D. Articles by Augustine, B.

 Previous Article  |  Next Article 

Journal of Bacteriology, January 2008, p. 555-563, Vol. 190, No. 2
0021-9193/08/$08.00+0     doi:10.1128/JB.01668-07
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

PhaP Is Involved in the Formation of a Network on the Surface of Polyhydroxyalkanoate Inclusions in Cupriavidus necator H16

Douglas Dennis,^1* Vicki Sein,¹ Edgar Martinez,¹ and Brian Augustine²

Department of Integrated Natural Sciences, Arizona State University, Glendale, Arizona 85306,¹ Department of Chemistry and Biochemistry, James Madison University, Harrisonburg, Virginia 22807²

Received 15 October 2007/ Accepted 25 October 2007

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Polyhydroxyalkanoate (PHA) inclusions are polymeric storage inclusions formed in some bacterial species when carbon levels are high but levels of another essential nutrient, such as nitrogen, are low. Though much is known about PHA synthesis, little is known about inclusion structure. In this study, atomic force microscopy (AFM) was employed to elucidate the structure of PHA inclusions at the nanoscale level, including the characterization of different layers of structure. AFM data suggest that underneath the inclusion envelope, there is a 2- to 4-nm-thick network layer that resides on top of a harder layer that is likely to be a crystalline lamellar polymer. The network is comprised of 20-nm-wide linear segments and junctions that are typically formed by the joining of three to four of the linear segments. In some cases, 50-nm globular structures that are raised 1 to 2 nm above the network are present at the junctions. These globular structures always have a central pore that is 15 nm in diameter. To determine if the major surface protein of PHA inclusions, PhaP, is involved in the structure of this network, inclusions from Cupriavidus necator H16 phaP were examined. No network structure was detected. Instead, apparently random globular structures were found on the surfaces of the inclusions. When PhaP levels were reconstituted in this strain by the addition of phaP on a plasmid, the network was also reconstituted, albeit in a slightly different arrangement from that of the wild-type network. We conclude that PhaP participates in the formation of the inclusion network.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Polyhydroxyalkanoates (PHAs) are bacterial storage polymers that accumulate when carbon levels are high but levels of another necessary nutrient, such as nitrogen, are low. In Cupriavidus necator (formerly Wautersia eutropha), PHA synthesis occurs in multiple inclusions that are generally 200 to 1,000 nm in size (2). The inclusions contain approximately 97.5% PHA, 2% protein, and 0.5% lipid (10), although some estimates of the lipid content are higher (3). PhaP is the dominant protein species on the inclusion and, under accumulation conditions, can constitute as much as 5% of the total cell protein. Because PhaP is found at the surface of the inclusion, it is thought to be the major structural protein of the inclusion (23, 34).

The precise nature of the inclusion surface has yet to be conclusively determined. Electron microscopy (EM) studies reported more than 40 years ago supported the existence of a membrane layer at the surface of inclusions isolated from Bacillus megaterium (18). Nine years later, freeze fracture experiments expanded these results by suggesting an arrangement in which there is a membranous outer layer, a solid layer underneath it, and an inner core of soft, deformable material (7). The finding of a solid second layer has been supported by studies that demonstrated a solid, crystalline lamellar shell at or near the surface of the inclusion with a soft, noncrystalline, hydrophobic core (17). In contrast to the membrane-solid layer-amorphous core inclusion model, later studies of inclusions from Pseudomonas species imaged a regular lattice-like surface architecture at the inclusion surface that is reminiscent of a bacterial S-layer (28). Because of this, some researchers suggested that instead of an outer phospholipid layer, a protein layer, most logically composed of PhaP with embedded synthesis and depolymerizing enzymes, covers the inclusion (14, 27). For a period of time, the difference between these two models seemed to be somewhat clarified by the determination by researchers that the outermost layer is 4 nm thick (21). This implies that it cannot be a lipid bilayer, which has been measured at 7 to 13 nm for C. necator H16 (21). Similarly, it does not seem possible that it is a monolayer of PhaP, because quantitative Western blot analyses have determined that there is only enough PhaP present to cover 27 to 54% of the total inclusion surface (31, 32). With these two possibilities seemingly ruled out, a newer model of the PHA inclusion structure was proposed in which the outer layer of the inclusion consists of a phospholipid monolayer, thus conforming to the 4-nm thickness previously determined for the outermost layer (29). This model, however, has been recently challenged by EM data showing that the outer boundary layers of PHA inclusions in Caryophanon latum were found to be 14 nm thick (13). Therefore, the nature of the surface of PHA inclusions and the mode of biogenesis of these inclusions remain undefined. In an attempt to resolve the basic structure of the PHA inclusion surface, we have employed high-resolution scanning probe microscopy. Our results suggest a model of inclusion structure that is consistent with the findings of the initial EM studies in that we have imaged an envelope that covers a solid crystalline shell. In addition, we describe a heretofore uncharacterized network layer residing between the envelope and the solid, crystalline lamellar layer. Preliminary studies suggest that this network is at least partially composed of PhaP.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Bacterial growth. Cupriavidus necator H16 was grown overnight in a nutrient broth culture, and this culture was used to inoculate a 50-ml culture (in a 250-ml flask) such that the initial optical density at 600 nm was between 0.05 and 0.1. A supplemented minimal medium was used (24). The constituents were (per liter) 9.0 g Na₂HPO₄·12 H₂O, 1.5 g KH₂PO₄, 0.1 g NH₄Cl, 0.2 g MgSO₄·7 H₂O, 0.02 g CaCl₂·H₂O, 1.2 mg Fe(III)NH₄-citrate, 0.10 ml SL6 trace elements (22), 5 ml 20% (wt/vol) Casamino Acids, 0.1 ml 0.5% (wt/vol) thiamine, and 0.4% (wt/vol) fructose. When cultures were incubated longer than 18 h, another aliquot of 0.4% fructose was added at approximately 20 h postinoculation. The culture was incubated at 30°C at a speed of 300 rpm. Cultures were normally grown to a final optical density at 600 nm of between 10 and 20.

Preparation of PHA inclusions. To minimize contamination during the isolation process, very small amounts of bacterial cells in relatively large volumes of buffer were subjected to sonication. Fifty microliters of cell culture was added to 1.5 ml of TN buffer (10 mM Tris [pH 7.5]-150 mM NaCl) in a microcentrifuge tube, vortexed, and pelleted (1 min at full speed). The pellet from this wash was resuspended in the 50 µl of TN buffer. In general, we added 5 µl of this cell suspension to 5 ml of buffer, but this was sometimes altered to obtain higher or lower inclusion densities. An aliquot of cell suspension was added to 5 ml of TN buffer and subjected to sonication using a Tekmar model TMX400 sonic disruptor with a tapered probe (5 mm at the tip) located about 4 mm from the bottom of a 13- by 100-mm sterile plastic culture tube (containing the cell suspension) in an ice bath. Sonication times and intensities varied depending on the level of inclusion deterioration that was desired. For gentle lysis, a sonication setting of 30% power for 60 s was employed. For a harsher sonication, in which the majority of the cells were lysed, a setting of 50% power for 5 min was used. To obtain the most disruptive sonication, the cells were subjected to sonication at 60% power for 5 to 10 min. Pulse times were 3.3 s with a resting phase of 9.9 s. The degree of cell lysis and inclusion disruption varied from culture to culture (probably due to the culture age, etc.), but if over- or undersonication occurred, the procedure was redone to obtain the type of lysis desired. Immediately after sonication, the entire contents of the tube were added to a fritted filter holder (25 mm) containing a 25-mm polycarbonate filter with 200-nm pores (Osmonics; KO2CPO2500). The sample was filtered quickly and washed with 5 to 15 ml of deionized water. The filter was immediately removed, fixed to a metal puck, and imaged.

Construction of C. necator H16 phaP. Forward primer delphaPF (CGAGCTCAGCGAGCATGGAAGTGGCCT), containing a SacI site at the 5' end (indicated by underlining), and reverse primer delphaPR (CTGTGCGCATCGGAGAATAGG) were employed to amplify a segment of phaP from nucleotides 395 to 1641 of the published sequence (GenBank accession number AF310426). The amplified segment was digested with SacI and Acc65I (an Acc65I site is located at base 1488) and ligated into digested vector pBluescript II SK+ to make pSK/phaP. Inverse PCR was performed on this vector using primers invphaPF (CGGATCCGAGGATCATTGCTGGTCTCC) and invphaPR (CGGATCCGCCTGATAACTGCCTGCGTTG). Both primers contain a BamHI site at their 5' ends. The forward primer initiates in a direction opposite to the transcriptional direction at base 769, which is at the third amino acid of PhaP. The reverse primer initiates in the opposite direction (the direction of transcription) at base 1334, which includes the final two amino acids of PhaP. The amplified fragment was digested with BamHI, ligated, and inserted (by electroporation) into Escherichia coli DH5 to make pSK/phaP. The truncated gene then consisted of an intact native promoter, the first three amino acids of PhaP, a BamHI site, the final two amino acids of PhaP, and downstream sequences. This was excised with SacI and Acc65I and ligated into digested pDM197 (8), followed by electroporation into E. coli S17-1 pir⁺ (pDM197/phaP). This strain, carrying the truncated gene, was mated with C. necator H16, and C. necator H16 phaP was selected using previously prescribed methods (8). The strain used was confirmed by PCR analysis, followed by DNA sequence analysis of the amplified PCR product.

Construction of C. necator H16 phaP(pBBR/phaP). The technique used to construct C. necator H16 phaP(pBBR/phaP) was the same as that described for the first PCR step above (amplification of bases 395 to 1641 of phaP), except that the SacI/Acc65I-cut fragment was ligated into digested pBBR1MCS-3 (15) and introduced into E. coli S17-1 (26) by electroporation. The plasmid, pBRR/phaP, was transferred to C. necator H16 phaP by conjugation.

Atomic force microscopy (AFM) imaging. A Digital Instruments multimode scanning probe microscope with a Nanoscope IIIa controller was used to image the inclusions in tapping mode. Applied Nanosytems ACT Si probes were used for cantilevers with a nominal length of 125 µm and a resonance frequency of 200 to 400 kHz. Both height and phase images were collected, but only the phase images were optimized. An "E" scanner, having a maximum scan range of 14 µm, was used. Scans were generally taken at a 512- by 512-line resolution with a scan velocity of 1 to 5 µm/s. Integral gain, proportional gain, amplitude set point, and drive amplitude were modulated as necessary to obtain optimum images. For very-high-resolution images (indicated in the figure legends), ACT-SSS (super-sharp-silicon) tips, having a guaranteed probe radius of curvature of less than 5 nm, were employed.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
By employing AFM analyses of inclusions released from C. necator H16 cells by sonication, we typically observe three types of inclusion surface structures (Fig. 1A to C): enveloped, networked, and naked. The proportion of each of these types within an inclusion population can be regulated by altering the degree of sonication. By the use of relatively gentle sonication conditions, a population of inclusions that is almost entirely enveloped is observed. If the severity of sonication is increased, this distribution changes such that both enveloped and networked inclusions are observed. If harsh sonication conditions are used (higher intensity and longer times), a population of inclusions that is almost entirely naked is observed. We have interpreted these data to mean that layers of the inclusion are being successively stripped from the inclusion surface as the sonication protocols become more rigorous. Therefore, we suggest that the enveloped inclusions comprise the intact inclusion surface and the envelope, the networked inclusions consist of the layer beneath the envelope, and the naked inclusions are inclusions from which both the envelope and the network layer have been stripped and represent the polymer core that is crystallized at its surface due to interaction of the hydrophobic polymer with air (17). This interpretation is supported by cross-sectional analysis of the three inclusion types which shows that the enveloped and networked inclusions maintain round shapes but that the naked inclusions have settled into energetically more stable conical shapes (Fig. 1D). Of particular interest to us is the network layer because it has not yet been described and because of its physical resemblance to eukaryotic cytoskeletons. Therefore, further high-resolution AFM analyses were conducted.

View larger version (147K): [in this window] [in a new window]

FIG. 1. AFM phase images of three types of inclusions generally imaged from cell lysates. (A) Enveloped inclusions. It is likely in this image that the cell made contact with the filter and subsequently lysed, leaving the inclusions arranged as if they were still in the cell. Original scan = 1.6 µm2. (B) Networked inclusions. The network is seen as lighter lines traversing the surface of the two inclusions. Gaps are seen as slightly darker areas. The terracing at the bottom of the inclusions is due to the probe not making sufficient contact with the inclusions as it tracked down their sides and should be disregarded. Original scan = 798 nm2. ACT-SSS probe used to acquire the image. (C) Naked inclusions. Note the light lines perpendicular to the perimeter pointing toward the center of the inclusion. These have no significant height signature. Also note the conical shape. Original scan = 1.3 µm2. (D) Display of cross-sectional analysis of typical inclusion types showing how inclusions differ in shape. Scale bars = 500 nm. AFM phase images detect the physical differences in a sample (19).

In preliminary studies, it was determined that AFM cantilevers with large tip radii (10 to 25 nm) tended to increase the perceived diameters of the linear segments of the network (tip convolution) (1, 33). In some cases, the spreading was so pronounced that the network looked less like a network and more like the surface that is observed on a human brain, and the linear segments were similar to the nanostructures recently reported by Kuchta et al. (16). For this reason, AFM probes that are very uniform and very sharp were utilized, with some of them having a radius of curvature as low as 2 nm. Although they are difficult to work with on convex soft objects, when properly used, these cantilevers generate images of outstanding resolution and clarity.

In scans ranging from about 1 µm² to 8 µm² in size, networked inclusions can typically be identified by the variegation of their surfaces in contrast to the lack of variegation of enveloped and naked inclusion surfaces (Fig. 1A to C). However, little detail about surface structure can be determined at this scan size (Fig. 2A). When scan sizes below 1 µm² are used, specific surface structures begin to be observed (Fig. 2B), but their interpretation is made possible only by having imaged and interpreted higher-resolution scans. With the highest-resolution scans, typically in the range of 150 nm² to 350 nm², the components of the network become interpretable. The network comprises a series of interconnected linear segments (Fig. 2C). In the highest-resolution images acquired, it appears that the linear segments may be composed of subunits, because faint lines perpendicular to the axes of the segments are observed, making it look like the subunits are "stacked" on top of each other to make the segments. Normally, the junctions of the linear segments are composed of three to four segments. The widths of the linear segments vary considerably, with an average width being 20 nm. This lateral measurement is subject to the error inherent in the radius of curvature of the cantilever tip (tip convolution) (1, 33), which should be minimal in this study because of the very fine tips used (2-nm radius of curvature). A structure that is often seen at the junction of the linear segments is a globular structure that is in the range of 40 to 70 nm in diameter, with the average being 50 nm (Fig. 2C). These globular structures always have a central pore with a very uniform diameter of 15 nm (Fig. 2D). The globular structures are usually 1 to 2 nm higher than the network.

View larger version (193K): [in this window] [in a new window]

FIG. 2. High-resolution AFM images of networked inclusions. An ACT-SSS probe was used to acquire these images. (A) Phase image of two inclusions with a network. The network is seen as the lighter, interconnected lines. Original scan = 798 nm2. Scale bar = 150 nm. (B) Phase image of one of the inclusions in panel A. At this resolution, the network appears more diffuse but can still be seen. Parallel striations in the gaps of the network (black arrows) are likely to be the crystalline lamellar polymer layer. Globular structures with central pores/depressions can be seen (white arrow). Original scan = 427 nm2. Scale bar = 150 nm. (C) Phase image of a portion of the network. In this image, a portion of the network was found detached and lying on the filter next to the inclusion. Some of the linear segments have faint cross striations. An 50-nm globular structure is indicated (white arrow). Original scan = 434 nm2. Scale bar = 75 nm. (D) Cross-section analysis of the globular structure and pore in panel C. The tip of the probe does not reach the bottom of the pore because the conical shape of the tip prevents this from happening. Note that the collar of the globular structure is raised above the rest of the network. The cross-section contour line starts at the circle in the image and at the left side of the graph.

Cross-sectional analysis of individual network linear segments reveals that they range in height from 2 to 4 nm (Fig. 3A and B). It is not known at this time whether the network strands reside on the surface of the crystalline layer beneath it or whether they are partially embedded in this layer. In high-resolution AFM height scans, pores were observed in the inclusion surface between the network strands. These pores were approximately 15 nm in diameter (Fig. 3A; cross-sectional data not shown). The pores could also be detected using AFM phase imaging. In addition to the pores, faint parallel striations could often be seen in the gaps of the network in phase images (Fig. 3C). The spacing of the striations was measured at 5 to 6 nm (Fig. 3D).

View larger version (142K): [in this window] [in a new window]

FIG. 3. Cross-section analyses. (A) AFM height image of network strands on an inclusion surface. In AFM height images, higher objects are lighter, and lower objects are darker. A cross-section line is drawn across several strands, which are numbered and correspond to the peaks in panel B. Pores are seen as dark circles about 15 nm in diameter. Original scan = 201 nm2. Scale bar = 100 nm. (B) Height contour of the cross-section line in panel A. The contour line begins on the left (at the cross hair in panel A) and moves to the right. (C) AFM phase image showing gaps in the network and putative crystallized polymer strands. The strands are seen as faint striations (black arrow). The pore can also be seen (white arrow). Original scan = 325 nm2. Scale bar = 100 nm. (D) Cross-section analyses of putative crystallized strands. The cross-section graph is of a single strand. Striations cannot be seen in the network gaps in panel A because their height signatures are not large enough to visualize changes with the z range used (50 nm).

To determine if the loss of PhaP, the dominant surface protein on PHA inclusions, causes a structural change in the surface network, a phaP strain was made in which the altered PhaP contained only its first three and last two amino acids with a GGATCC linker in between. This strain, C. necator H16 phaP, grew more slowly than the wild-type strain and synthesized larger inclusions, generally one per cell. Inclusions from C. necator H16 phaP were much less uniform in shape than those of the wild type (Fig. 4A and B). They often had irregularly shaped perimeters. In addition, they exhibited no linear network structures. Instead, a large accumulation of globular structures was normally present on the surface (Fig. 4A and B). The number of globular structures was linked to the degree of sonication, with inclusions from more rigorously sonicated cells having little in the way of globular structure. Occasionally, globular structures with central pores were tentatively identified on an inclusion (Fig. 4C and D), but given the randomness of the surface structure, it was difficult to positively identify these as the same globular structures seen in the networked inclusions. The 5- to 6-nm parallel striations were also observed (data not shown) but less frequently, because the inclusions tended to be almost completely covered with globular material. These large inclusions were more difficult to image than inclusions from wild-type cells. Often the cantilevers gouged the inclusions at their edges and/or punctured the inclusions as they traversed their surfaces. We are unsure at this time whether this is due to the increased height signature and the inherent difficulties in traversing it or to an increase in fragility of the inclusion surface.

View larger version (149K): [in this window] [in a new window]

FIG. 4. AFM analysis of inclusions from C. necator H16 phaP that have had their envelopes removed. (A and B) Phase images of inclusions. Little in the way of regular surface structure can be identified. In panel B, the presence of a tentative, 50-nm-diameter globular structure is indicated by an arrow. Original scans, 1.3 µm2 (A) and 996 nm2 (B). Scale bars = 100 nm. (C and D) The putative globular structure is analyzed by cross-sectional analysis (white line traversing structure in panel C). The pore is approximately 15 nm in diameter, and the entire structure is 62 nm in diameter. The structure is raised 2.5 nm above the surface of the inclusion. Bar = 100 nm.

When plasmid-borne phaP under the control of its own promoter was introduced into C. necator H16 phaP, the strain regained a normal growth rate and PHA accumulation levels (data not shown). The inclusions, however, were much smaller. The network was restored to the surface of the inclusion (Fig. 5A and B), but it appeared to differ slightly from the inclusion network observed in C. necator H16 in that the linear segments were generally 5 to 10 nm wider, giving the appearance of a much more robust structure. It did not appear that there were as many junctions in the network, though this finding is somewhat subjective. The heights of the individual strands were the same as those found in inclusions from wild-type cells (cross-sectional data not shown). Pores that were 15 nm in diameter were observed, but they were not as often located in distinct globular structures of 50 nm in diameter (Fig. 5B). Instead, they were at the junctions of the widened network strands.

View larger version (131K): [in this window] [in a new window]

FIG. 5. AFM phase images of inclusions from C. necator H16 phaP(pBBR/phaP). (A) Original scan = 1.8 µm2. Scale bar = 200 nm. (B) Enlargement of panel A showing the network on a single inclusion and an 15-nm pore in a widened area of the network (white arrow). Scale bar = 200 nm.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
AFM offers an extremely powerful tool for the characterization of biological surfaces (6). Not only can the technique be used to image structures at the nanoscale level, the sample preparation required is much less severe, and thus, delicate structures can be imaged that might otherwise be compromised using traditional sample preparation techniques required for EM studies. In fact, from the time of harvesting bacterial cells from the culture, high-resolution images could be obtained within 15 min.

The AFM data support the early EM studies in that they confirm the existence of an envelope with a crystalline lamellar shell beneath it (7, 17, 18). The AFM data differ significantly from those of prior studies, however, in that AFM analysis has identified a network layer residing between the envelope and the crystalline lamellar shell. The structure of this network is somewhat reminiscent of the structure of the red blood cell cytoskeleton (Fig. 6) (30). PhaP is likely to be a component of the network because when phaP is deleted, the network is not formed, and when phaP is inserted on a low-copy-number plasmid into C. necator H16 phaP, the network is again formed.

View larger version (198K): [in this window] [in a new window]

FIG. 6. AFM images of an inclusion network (A, C, and D) and an erythrocyte cytoskeleton (B). Scale bars = 100 nm. (A) Phase image of a network. In this instance, the network was found detached from an inclusion, lying on the filter substrate. This allowed imaging at a very high resolution. A putative, 50-nm-diameter globular structure with a central pore is shown (white arrow). Original scan = 434 nm2. (B) Erythrocyte skeleton imaged in tapping mode in air. Note the difference in scale between this image and the inclusion images. Taken from reference 30 with permission from the publisher. (C) Surface plot of panel A. This AFM tool renders the image more three-dimensional by tilting the image and varying the light angle. Two putative, 50-nm-diameter globular structures are pointed out (white arrows). Others are present, however. (D) Height image of panel C. z range = 50 nm.

It could be argued that PhaP is a regulatory protein that stimulates the production of another protein that composes the network, but this ignores the fact that PhaP is the dominant protein located on the inclusion surface. Furthermore, the coverage of the network is consistent with prior studies that showed that PhaP covers 27 to 54% of the inclusion surface (31, 32) and with the fact that alteration of the PhaP levels alters the size of the inclusion (34), suggesting a structural function. A structural function is also suggested for the imaged network by the fact that removal of the network results in a conformation change in the inclusion from rounded to conical. Finally, the calculated radius of PhaP, 2.1 nm, is consistent with the height of the network (32).

The inclusion network that is formed in strains where phaP is located on a low-copy-number plasmid appears to be slightly different than the network formed on inclusions in wild-type cells. The linear portions are 5 to 10 nm wider, and the network appears to have longer linear runs between junctions (though we have not mathematically confirmed this last observation). This may be due to an increase in intracellular PhaP levels mediated by a gene dosage effect of the plasmid-borne phaP gene. It may also be that other proteins participate in the network, especially the junctions/globular structures, and that when the ratio of PhaP to these other proteins is changed, it changes the structure of the network. We are currently focusing on PhaR, an inclusion surface protein that is known to interact with PhaP, to determine if changing the PhaP/PhaR ratios alters the arrangement of the network. Experimental support for this notion comes from research on recombinant E. coli harboring the C. necator H16 biosynthesis genes (phaCAB), in which the addition of phaP, phaR, or both resulted in changes in size and shape of the formed inclusions (25).

It is probable that "naked" inclusions are inclusions that have been stripped to their crystalline lamellar shell, because parallel striations that conform to the previously measured width of PHA fibrils (9) are present. These striations can also be imaged in the gaps of the network, suggesting that the network resides on top of the crystalline lamellar shell. Interestingly, 15-nm-diameter pores were observed in the crystalline lamellar shell. Though their function is unknown, it is logical to assume that these pores provide a portal to the amorphous polymer core residing beneath the crystalline shell, both for synthesis and for depolymerization.

The pores identified in the globular structures are also intriguing. The globular structures themselves possess a fairly broad range of size, yet the pores always occur centrally, singly, and with diameters very close to 15 nm. This uniformity has structural implications. Pores of the same size have also been identified in the envelope layer (unpublished data). One might suppose that in the intact inclusion, the 15-nm pores found in all three layers are aligned, thus providing a portal of ingress and egress through the three layers. In fact, if the pores in all three layers align, it seems likely that the pores are the sites of PHA metabolism, because there is no other plausible way that molecules could enter the inclusion fast enough to sustain the high rate of polymerization that occurs during PHA accumulation.

It is not a certainty that the networks that are imaged in this study are representative of the network in an intact inclusion. It may be that when the envelope is removed, portions of the network are lost either through physical trauma or because they are attached to the envelope. It does appear that the network is fairly easily removed from the inclusion surface, because the sum total of the AFM images acquired reveal the network in many stages of deterioration (data not shown). In some cases, only a few linear remnants are present on top of the crystalline lamellar shell. In fact, inclusions with partially degraded networks have assisted in the interpretation of the three-layer inclusion hypothesis. In this paper, only images of the most complete network are displayed.

With respect to network deterioration, it is possible that the globular structures are easily lost from the network, because in an earlier study in which inclusions resting on filters were treated with a dilute sodium dodecyl sulfate solution, many 50-nm globular structures were imaged lying on the filter beside the inclusion but not in other areas of the filter (4). Also, if there is a 1:1:1 relationship between the pores in the crystalline lamellar layer, the pores in the globular structures, and the pores in the envelope, then there is a deficit in the number of globular structure pores. Additionally, there are many pores observed in the crystalline lamellar layer that do not have a corresponding globular structure over them. The strongest evidence for this supposition comes from the prior AFM study in which globular structures with central 15-nm pores were observed at a high density under the surfaces of enveloped inclusions (4).

An issue to be considered is the effect of drying on the network. It has been reported that drying of the erythrocyte membrane for tapping-mode AFM in air causes a slight deformation of the attached cytoskeleton (30). This is a possibility for PHA inclusions, but because the network is apparently linked to the crystalline layer (otherwise it would be totally lost during sonication), it us unlikely that it deforms much beyond what occurs as the polymer crystallizes. Once crystallized, the surfaces of inclusions are quite stable, and they have been imaged over long periods of time with no apparent alteration in their structures. To determine what effect, if any, crystallization of the polymer might have on the arrangement of the network, AFM liquid-cell studies are now being conducted.

An obvious question raised by this study is why the network has not been observed prior to this report. There are at least two reasons why this may have occurred. First, the network layer is only 2 to 4 nm thick, is not a solid layer, and is easily detached from the inclusion. It may be that during the preparation of the inclusions for electron microscopy, the network was lost, or perhaps it is simply not discernible in a cross section because of its incomplete nature. Or it may be that the layer is simply below the limits of detection for those particular EM protocols. This is not unique in biology. It is only recently that cryoelectron microscopy identified a 2- to 4-nm granular layer on the outer face of the cytoplasmic membrane in gram-positive bacteria (35) and a periplasmic space also in gram-positive organisms (20). The second explanation is that in prior studies, the envelope was removed by the EM preparation procedure, leaving only the network and the crystalline lamellar layer, which would constitute an electron-dense layer that is approximately 4 nm thick. Therefore, when previous studies reported the boundary layer as being 4 nm thick (21), it may be that the network/lamellar shell was being measured. It is well established that EM preparatory procedures are deleterious to membranes and related attached structures (5). It is noteworthy that recent EM studies have detected "regularly oriented" proteins on the surface of PHA inclusions from Caryophanon latum (13). However, the level of resolution was not high enough for a significant interpretation.

In future experimentation, we will continue to study the network layer and examine how alterations in growth conditions, bacterial species, and gene deletions alter its structure. We are also interested in the possibility that the network may provide a focal point where proteins can transiently associate to mediate cellular/PHA reactions. This is suggested by recent studies in which it has been shown that phasins (PhaP) perform this function for triacylglycerol inclusions (11). It is also of interest that the 25-amino-acid, alanine-rich carboxy terminus of PhaP exhibits minor homology to the mouse clathrin coat assembly protein (GenBank accession number Q61548), a membrane-embedded protein that can recruit and release clathrin and other proteins (12). In short, we hope to gather additional evidence that will support the existence of the network layer and elucidate its function from an inclusion and cellular perspective.

 
This work was supported by grants 0110232, 0071717, and 0520882 from the National Science Foundation.

 
* Corresponding author. Mailing address: Department of Integrated Natural Sciences, Arizona State University West, P. O. Box 37100, Phoenix, AZ 85069-7100. Phone: (602) 543-6934. Fax: (602) 543-6073. E-mail: Douglas.dennis{at}asu.edu

Published ahead of print on 2 November 2007.

Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 

1
1. Allen, M. J., N. V. Hud, M. Balooch, R. J. Tench, W. J. Siekhaus, and R. Balhorn. 1992. Tip-radius-induced artifacts in AFM images of protamine-complexed DNA fibers. Ultramicroscopy 42-44:1095-1100.
2
2. Anderson, A. J., and E. A. Dawes. 1990. Occurrence, metabolism, metabolic role, and industrial uses of bacterial polyhydroxyalkanoates. Microbiol. Rev. 54:450-472.[Abstract/Free Full Text]
3
3. Anderson, A. J., G. W. Haywood, and E. A. Dawes. 1990. Biosynthesis and composition of bacterial poly(hydroxyalkanoates). Int. J. Biol. Macromol. 12:102-105.[CrossRef][Medline]
4
4. Dennis, D., C. Liebig, T. Holley, K. S. Thomas, A. Khosla, D. Wilson, and B. Augustine. 2003. Preliminary analysis of polyhydroxyalkanoate inclusions using atomic force microscopy. FEMS Microbiol. Lett. 226:113-119.[CrossRef][Medline]
5
5. Dubochet, J., A. W. McDowall, B. Menge, E. N. Schmid, and K. G. Lickfeld. 1983. Electron microscopy of frozen-hydrated bacteria. J. Bacteriol. 155:381-390.[Abstract/Free Full Text]
6
6. Dufrêne, Y. F. 2004. Refining our perception of bacterial surfaces with the atomic force microscope. J. Bacteriol. 186:3283-3285.[Free Full Text]
7
7. Dunlop, W. F., and A. W. Robards. 1973. Ultrastructural study of poly-β-hydroxybutyrate granules from Bacillus cereus. J. Bacteriol. 114:1271-1280.[Abstract/Free Full Text]
8
8. Edwards, R. A., L. H. Keller, and D. M. Schifferli. 1998. Improved allelic exchange vectors and their use to analyze 987P fimbria gene expression. Gene 207:149-157.[CrossRef][Medline]
9
9. Ellar, D., D. G. Lundgren, K. Okamura, and R. H. Marchessault. 1968. Morphology of poly-beta-hydroxybutyrate granules. J. Mol. Biol. 35:489-502.[CrossRef][Medline]
10
10. Griebel, R., Z. Smith, and J. M. Merrick. 1968. Metabolism of poly-beta-hydroxybutyrate. I. Purification, composition, and properties of native poly-beta-hydroxybutyrate granules from Bacillus megaterium. Biochemistry 7:3676-3681.[CrossRef][Medline]
11
11. Hanisch, J., M. Waltermann, H. Robenek, and A. Steinbuchel. 2006. The Ralstonia eutropha H16 phasin PhaP1 is targeted to intracellular triacylglycerol inclusions in Rhodococcus opacus PD630 and Mycobacterium smegmatis mc2155, and provides an anchor to target other proteins. Microbiology 152:3271-3280.[Abstract/Free Full Text]
12
12. Hanley, S. Z., D. J. Pappin, D. Rahman, A. J. White, K. M. Elborough, and A. R. Slabas. 1999. Re-evaluation of the primary structure of Ralstonia eutropha phasin and implications for polyhydroxyalkanoic acid granule binding. FEBS Lett. 447:99-105.[CrossRef][Medline]
13
13. Jendrossek, D., O. Selchow, and M. Hoppert. 2007. Poly(3-hydroxybutyrate) granules at the early stages of formation are localized close to the cytoplasmic membrane in Caryophanon latum. Appl. Environ. Microbiol. 73:586-593.[Abstract/Free Full Text]
14
14. Jurasek, L., and R. H. Marchessault. 2002. The role of phasins in the morphogenesis of poly(3-hydroxybutyrate) granules. Biomacromolecules 3:256-261.[CrossRef][Medline]
15
15. Kovach, M. E., P. H. Elzer, D. S. Hill, G. T. Robertson, M. A. Farris, R. M. Roop II, and K. M. Peterson. 1995. Four new derivatives of the broad-host-range cloning vector pBBR1MCS, carrying different antibiotic-resistance cassettes. Gene 166:175-176.[CrossRef][Medline]
16
16. Kuchta, K., L. Chi, H. Fuchs, M. Potter, and A. Steinbuchel. 2007. Studies on the influence of phasins on accumulation and degradation of PHB and nanostructure of PHB granules in Ralstonia eutropha H16. Biomacromolecules 8:657-662.[CrossRef][Medline]
17
17. Lauzier, C., R. H. Marchessault, P. Smith, and H. Chanzy. 1992. Structural study of isolated poly(beta-hydroxybutyrate) granules. Polymer 33:823-827.[CrossRef]
18
18. Lundgren, D. G., R. M. Pfister, and J. M. Merrick. 1964. Structure of poly-beta-hydroxybutyric acid granules. J. Gen. Microbiol. 34:441-446.[Abstract/Free Full Text]
19
19. Magonov, S. N., V. Elings, and M.-H. Whangbo. 1997. Phase imaging and stiffness in tapping-mode atomic force microscopy. Surface Science 375:L385-L391.[CrossRef]
20
20. Matias, V. R. F., and T. J. Beveridge. 2006. Native cell wall organization shown by cryo-electron microscopy confirms the existence of a periplasmic space in Staphylococcus aureus. J. Bacteriol. 188:1011-1021.[Abstract/Free Full Text]
21
21. Mayer, F., and M. Hoppert. 1997. Determination of the thickness of the boundary layer surrounding bacterial PHA inclusion bodies and implications for models describing the molecular architecture of this layer. J. Basic Microbiol. 37:45-52.[CrossRef]
22
22. Pfennig, N., and K. D. Lippert. 1966. Uber das Vitamin B-12-bedurfnis Phototopher Schwefelbakteriein. Arch. Mikrobiol. 55:245.[CrossRef]
23
23. Pieper-Fürst, U., M. H. Madkour, F. Mayer, and A. Steinbüchel. 1995. Identification of the region of a 14-kilodalton protein of Rhodococcus ruber that is responsible for the binding of this phasin to polyhydroxyalkanoic acid granules. J. Bacteriol. 177:2513-2523.[Abstract/Free Full Text]
24
24. Schlegel, H. G., H. Kaltwasser, and G. Gottschalk. 1961. A submersion method for culture of hydrogen-oxidizing bacteria: growth physiological studies. Arch. Mikrobiol. 38:209-222.[CrossRef][Medline]
25
25. Seo, M. C., H. D. Shin, and Y. H. Lee. 2003. Functional role of granule-associated genes, phaP and phaR, in poly-beta-hydroxybutyrate biosynthesis in recombinant E. coli harboring phbCAB operon. Biotechnol. Lett. 25:1243-1249.[CrossRef][Medline]
26
26. Simon, R., U. Priefer, and A. Pühler. 1983. A broad host range mobilization system for in vivo genetic engineering: transposon mutagenesis in gram negative bacteria. Bio/Technology 1:784-791.[CrossRef]
27
27. Steinbuchel, A., K. Aerts, W. Babel, C. Follner, M. Liebergesell, M. H. Madkour, F. Mayer, U. Pieper-Furst, A. Pries, and H. E. Valentin. 1995. Considerations on the structure and biochemistry of bacterial polyhydroxyalkanoic acid inclusions. Can. J. Microbiol. 41(Suppl. 1):94-105.[Medline]
28
28. Stuart, E. S., R. W. Lenz, and R. C. Fuller. 1995. The ordered macromolecular surface of polyester inclusion bodies in Pseudomonas oleovorans. Can. J. Microbiol. 41:84-93.[Medline]
29
29. Stubbe, J., and J. Tian. 2003. Polyhydroxyalkanoate (PHA) homeostasis: the role of PHA synthase. Nat. Prod. Rep. 20:445-457.[CrossRef][Medline]
30
30. Swihart, A. H., J. M. Mikrut, J. B. Ketterson, and R. C. Macdonald. 2001. Atomic force microscopy of the erythrocyte membrane skeleton. J. Microsc. 204:212-225.[Medline]
31
31. Tian, J., A. J. Sinskey, and J. Stubbe. 2005. Kinetic studies of polyhydroxybutyrate granule formation in Wautersia eutropha H16 by transmission electron microscopy. J. Bacteriol. 187:3814-3824.[Abstract/Free Full Text]
32
32. Tian, J., A. He, A. G. Lawrence, P. Liu, N. Watson, A. J. Sinskey, and J. Stubbe. 2005. Analysis of transient polyhydroxybutyrate production in Wautersia eutropha H16 by quantitative Western analysis and transmission electron microscopy. J. Bacteriol. 187:3825-3832.[Abstract/Free Full Text]
33
33. Velegol, S. B., S. Pardi, X. Li, D. Velegol, and B. E. Logan. 2003. AFM imaging artifacts due to bacterial cell height and AFM tip geometry. Langmuir 19:851-857.[CrossRef]
34
34. Wieczorek, R., A. Pries, A. Steinbüchel, and F. Mayer. 1995. Analysis of a 24-kilodalton protein associated with the polyhydroxyalkanoic acid granules in Alcaligenes eutrophus. J. Bacteriol. 177:2425-2435.[Abstract/Free Full Text]
35
35. Zuber, B., M. Haenni, T. Ribeiro, K. Minnig, F. Lopes, P. Moreillon, and J. Dubochet. 2006. Granular layer in the periplasmic space of gram-positive bacteria and fine structures of Enterococcus gallinarum and Streptococcus gordonii septa revealed by cryo-electron microscopy of vitreous sections. J. Bacteriol. 188:6652-6660.[Abstract/Free Full Text]

---

Journal of Bacteriology, January 2008, p. 555-563, Vol. 190, No. 2
0021-9193/08/$08.00+0     doi:10.1128/JB.01668-07
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

This article has been cited by other articles:

• Jendrossek, D. (2009). Polyhydroxyalkanoate Granules Are Complex Subcellular Organelles (Carbonosomes). J. Bacteriol. 191: 3195-3202 [Full Text]  

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Dennis, D. Articles by Augustine, B. Search for Related Content PubMed PubMed Citation Articles by Dennis, D. Articles by Augustine, B.