Here we present a protocol to produce gram-negative Escherichia coli (E. coli) spheroplasts and gram-positive Bacillus megaterium (B. megaterium) protoplasts to clearly visualize and rapidly characterize peptide-bacteria interactions. This provides a systematic method to define membrane localizing and translocating peptides.
The use of confocal microscopy as a method to assess peptide localization patterns within bacteria is commonly inhibited by the resolution limits of conventional light microscopes. As the resolution for a given microscope cannot be easily enhanced, we present protocols to transform the small rod-shaped gram-negative Escherichia coli (E. coli) and gram-positive Bacillus megaterium (B. megaterium) into larger, easily imaged spherical forms called spheroplasts or protoplasts. This transformation allows observers to rapidly and clearly determine whether peptides lodge themselves into the bacterial membrane (i.e., membrane localizing) or cross the membrane to enter the cell (i.e., translocating). With this approach, we also present a systematic method to characterize peptides as membrane localizing or translocating. While this method can be used for a variety of membrane-active peptides and bacterial strains, we demonstrate the utility of this protocol by observing the interaction of Buforin II P11A (BF2 P11A), an antimicrobial peptide (AMP), with E. coli spheroplasts and B. megaterium protoplasts.
Antimicrobial peptides (AMPs) have gained attention due to their potential use as alternatives to conventional antibiotics1,2,3,4,5. AMPs kill bacteria by either translocating across the cell membrane and interacting with intracellular components such as nucleic acids or by permeabilizing the membrane causing leakage of cell contents6. In addition to their use as antibiotics, translocating AMPs may be adapted for drug delivery applications because they can non-disruptively cross the impermeable cell membrane7,8. We, therefore, seek to understand fundamental AMP mechanisms of action to lay the foundation for their use in drug design.
Confocal microscopy offers a way to assess localization patterns of fluorescently labeled AMPs in bacterial cells providing insights into their mechanism of action9,10,11,12,13,14. By labeling the membrane of the bacteria, one can determine if a fluorescently labeled peptide localizes to the membrane or the intracellular space of a bacterial cell. However, this technique is limited by the small size and rod shape of bacteria, which can make imaging challenging due to the resolution limits of conventional light microscopes and the variable orientation of the bacteria on the slide15.
The goal of the presented method is to enable enhanced visualization of the fluorescently labeled peptide localization patterns using confocal microscopy. Visualization is enhanced by turning the small, thin, rod-shaped gram-negative Escherichia coli (E. coli) and gram-positive Bacillus megaterium (B. megaterium) bacteria into enlarged, spherical forms referred to as spheroplasts (for gram-negative strains) and protoplasts (for gram-positive strains)16,17,18,19,20,21. Spheroplasts and protoplasts are easier to image because of both their increased size and their symmetric shape, which makes the orientation of a bacterium on a slide irrelevant for its imaging. In addition, we present a systematic approach to quantitatively analyze confocal microscopy data in order to characterize AMPs as either membrane localizing or translocating. Applying these methods makes it easier to distinguish fluorescently labeled peptide localization patterns. The protocols presented here can be used to assess the localization of a variety of membrane-active agents other than AMPs, including cell-penetrating peptides.
One distinct advantage of this technique is that it provides insights into the mechanism of action of AMPs on a single cell level, which may reveal cell-to-cell heterogeneity15, as opposed to other fluorescence assays commonly used to identify the mechanisms of action of AMPs, which only provide bulk estimates9,22,23,24,25. The use of spheroplasts and protoplasts in order to assess AMP cell entry is particular useful26 because they are more physiologically relevant15 than other models used for assessing cell entry, such as lipid vesicles24.
1. Solution Preparation
NOTE: Prepare solutions described in steps 1.1–1.9 and 1.8–1.11 in order to produce E. coli spheroplasts and B. megaterium protoplasts, respectively.
2. Preparation of Overnight Culture
NOTE: Perform sections 2-4 using appropriate sterile techniques. If desired, a bacterial strain can contain a plasmid for antibiotic resistance to reduce the potential contamination. If using a strain with antibiotic resistance, add the necessary antibiotics in steps 2.1, 3.1–3.2, and 4.1–4.4.
3. Preparation of Gram-negative E. coli Spheroplasts
4. Preparation of Gram-positive B. megaterium Protoplasts
5. Preparation of Peptide Solution and Membrane Dye for Imaging
6. Visualization of E. coli Spheroplasts and B. megaterium Protoplasts Using Confocal Microscopy
7. Characterization of AMP Localization
By enlarging bacteria and making them spherical, we can easily distinguish whether peptides localize to the bacterial membrane or readily translocate across the bacterial membrane. The resolution limits of conventional light microscopes make it challenging to distinguish whether peptide signals arise from the membrane or intracellular space in normal bacteria because signals localized to the membrane will appear to overlap with the intracellular space (Figure 3A). In contrast, the enlarged size of spheroplasts (2–5 µm) and protoplasts (2–3 µm) compared to normal bacteria, which are typically only 1 µm in diameter, results in clear resolution between the membrane marker and the intracellular space making it easier to distinguish peptide localization (Figure 3B).
Here, spheroplasts and protoplasts are used to show the localization pattern of the N-terminal FITC labeled AMP BF2 P11A. We observe FITC labeled BF2 P11A to both localize to the membrane and translocate across the cell membrane of E. coli spheroplasts and B. megaterium protoplasts (Figure 4). While wild type BF2 has been observed to translocate across E. coli and lipid vesicle membranes, the P11A mutation is known to reduce this translocation13,27,28. We utilized a non-antibiotic resistant B. megaterium and ampicillin resistant Top 10 E. coli (pET45B) in data presented. The ampicillin resistant E. coli was grown in the presence of ampicillin (349.4 g/mol) at a final concentration in solution of 25 µg/mL. For imaging, following the systematic approach outlined in section 7, ROIs were drawn on the membrane, intracellular space, and the background (Figure 5B). The ratio of intracellular peptide fluorescence intensity to peptide fluorescence intensity on the membrane was calculated for each spheroplast or protoplast interacting with BF2 P11A using the equation described in section 7.2. Figure 5A shows the distribution of this ratio for all spheroplasts and protoplast scored. The spheroplasts and protoplasts scored largely fell into two groups, with the majority of spheroplasts or protoplasts having a ratio of less than 0.3 or greater than 1.
Given the distinct groups that the ratios fall into, we defined peptide localization as translocating when the ratio calculated in section 7.2 was greater than or equal to 1. Conversely, peptide localization was defined as membrane localizing when the ratio described in section 7.2 was less than 1. Following this method of scoring, we observed a similar localization pattern for BF2 P11A in E. coli spheroplasts and B. megaterium protoplasts with BF2 P11A localizing to the membrane in 71% and 70% of cases for E. coli spheroplasts and B. megaterium protoplasts, respectively (Table 1).
Figure 1: Representative images of E. coli. (A) E. coli bacteria (B) E. coli snake and (C) E. coli spheroplast. Images were taken at 100X magnification. Please click here to view a larger version of this figure.
Figure 2: Representative images of B. megaterium. (A) B. megaterium bacteria and (B) B. megaterium protoplast. Images were taken at 100X magnification. Please click here to view a larger version of this figure.
Figure 3: Representative images of B. megaterium. (A) B. megaterium bacterial cell and (B) B. megaterium protoplast labeled with the membrane marker di-8-ANEPPS. Images from the middle z-stack are shown at 100X (A) and 63X (B) magnification. Please click here to view a larger version of this figure.
Figure 4: Representative E. coli spheroplasts and B. megaterium protoplasts interacting with FITC labeled BF2 P11A. (A) E. coli spheroplast showing BF2 P11A localizing to the membrane. (B) E. coli spheroplast showing BF2 P11A translocating across the membrane. (C) B. megaterium protoplast showing BF2 P11A localizing to the membrane. (D) B. megaterium protoplast showing BF2 P11A translocating across the membrane. E. coli spheroplasts and B. megaterium protoplasts are labeled with the membrane marker di-8-ANEPPS (red) and FITC labeled BF2 P11A (green). Images from the middle z-stack are shown at 63X magnification. Please click here to view a larger version of this figure.
Figure 5: Analysis of peptide localization patterns. (A) Distribution of ratio of intracellular fluorescence intensity (ROI 2) to membrane fluorescence intensity (ROI 1) after background subtraction ((ROI 2-ROI 3)/(ROI 1-ROI 3)) for E. coli spheroplasts and B. megaterium protoplasts labeled with BF2 P11A (B) E. coli spheroplast interacting with FITC labeled BF2 P11A. Circular Regions of Interest (ROI) 0.3 µm in diameter drawn on the membrane (ROI 1) intracellular space (ROI 2) and background (ROI 3). The fluorescence intensity of peptide was quantified in each ROI and the background fluorescence was accounted for by subtracting the fluorescence intensity in ROI 3 from the fluorescence intensity in ROI 1 and ROI 2. Please click here to view a larger version of this figure.
Bacterial strain | No. spheroplast or protoplast | % Membrane localizing | % Translocating |
E. coli | 84 | 71 | 29 |
B. megaterium | 70 | 70 | 30 |
Table 1: The localization pattern of BF2 P11A interacting with E. coli spheroplasts and B. megaterium protoplasts.
The protocols presented here make it feasible for researchers to more rapidly obtain larger sample sizes of bacterial images because the enlarged, spherical bacteria are much easier to locate, orient, and image. This enhanced ability to collect data is valuable in several respects. First, it enables a more systematic quantitative analysis of peptide localization patterns. While qualitative trends can be demonstrated from smaller sets of images, only a large sample set of high-quality images reveal more nuanced trends in localization patterns, such as the percentage of cells where a peptide translocates versus localizes to the membrane. Also, the ability to better resolve the internal fluorescence from the membrane localization makes it easier to perform time course studies to assess how quickly peptides interact with bacterial cells as well as how localization patterns change over time. The improved resolution also allows researchers to consider trends in peptide localization that are not possible with smaller bacteria. For example, in some of our observed spheroplasts and protoplasts, peptides appear to localize to specific sections of the membrane, i.e., the peptide signals do not form a complete, cohesive ring around the bacteria and instead show some punctate labeling. In other cases, peptides may not be distributed entirely evenly inside the bacterial membrane, i.e., the peptide signals do not completely fill the inner space of the bacteria. While we have not pursued these trends in our current analysis, considering such localization trends becomes feasible when imaging spheroplasts and protoplasts instead of standard bacterial cells. These advantages would also apply to determining the localization of peptides other than AMPs, such as cell-penetrating peptides. Spheroplasts and protoplasts produced with these approaches could also be utilized for non-imaging experiments that benefit from increased cellular size, such as patch-clamp electrophysiology measurements18,19.
A laser scanning confocal microscope was utilized in the development of our imaging protocol, but it is critical to optimize parameters for each specific imaging system. This includes selecting a set of parameters that maximize the detection of peptide and membrane dye fluorescence, while minimizing bleed-through from the membrane dye into the peptide's emission channel, and from the FITC-labeled peptide into the membrane's emission channel. Bleed-through from the peptide into the membrane's emission channel was avoided by selecting a wavelength range for the membrane's emission channel that did not include any portion of FITC's emission spectrum. Bleed-through from the membrane dye, di-8-ANEPPS, into the peptide's emission channel was more difficult to avoid because di-8-ANEPPS' emission spectrum overlaps with the majority of FITC's emission spectrum. Specific to this protocol, bleed-through from the membrane dye into the peptide's emission channel can result in the false characterization of AMPs as membrane localized. One can determine if the bleed-through of this nature is occurring by imaging spheroplasts or protoplasts that are exclusively labeled with membrane dye and checking to see if the fluorescence is visible in the peptide's emission channel. Various imaging parameters, including the wavelength range used to define the peptide's emission channel and the gain and offset values, can be systematically tested to determine a set of parameters that minimizes bleed-through. Once a set of parameters is established, it is critical to evaluate whether imaging yields strong detection of fluorescence signal when spheroplasts or protoplasts labeled with both peptide and membrane dye are utilized. Through this approach, we successfully established imaging parameters that minimized the effect of the bleed-through while maximizing the fluorescence detection of both FITC-labeled AMP and the membrane dye di-8-ANEPPS. Specifically, we found that the emission bleed-through from the membrane dye into the peptide channel was minimized using emission wavelength ranges of 499–532 nm (peptide) and 670–745 nm (membrane), and when the gain was adjusted to be ≤800 volts (V) (peptide) and ≤900 V (membrane) and the offset was adjusted to ≤ -10.0% (peptide, membrane).
Although we focused on E. coli and B. megaterium, the presented protocols could be adapted to produce spheroplasts or protoplasts from many different bacterial strains. For example, we have been able to successfully produce Bacillus subtilis protoplasts that were 1–2 µm in diameter using the protocol outlined in section 4, and adaptations of the protocol could be made to further optimize size. The ability to observe peptide localization patterns in both gram-negative and gram-positive bacteria is particularly useful to the study of AMPs because differences in the cell wall structure between these two classes of bacteria may affect how AMPs interact with cellular membranes. However, many imaging studies of AMPs to date have solely focused on model E. coli strains and, thus, may not reflect the behavior of peptides with gram-positive bacterial strains.
Spheroplasts and protoplasts differ from normal bacteria, most notably in their lack of an outer cell wall, and consequently may not be good models for all experimental questions. For example, the outer membrane of gram-negative bacteria has been shown to act as a molecular sieve for larger molecules29. Larger peptides, therefore, may be able to translocate in spheroplasts when they would not do so in normal bacteria. Additionally, detailed studies of the interaction of AMP with the bacteria cell wall, outer membrane, and cytoplasmic membrane cannot be performed using spheroplasts and protoplasts. For example, recent work from the Weisshaar Lab has utilized imaging to note the more precise positions of the peptide in the CM15 and melittin mechanisms of action30,31. However, our approach, as described here, does not require the implementation of microfluidics into microscopy instrumentation in order to achieve the required resolution31. Although previous work has shown that spheroplasts are viable21,32, they nonetheless likely have some metabolic and physiological differences from "normal" bacteria that could potentially alter membrane interactions in some cases. Additionally, because we have not assessed the viability of spheroplast populations, one cannot distinguish between a peptide that is readily able to translocate across the cell membrane of a living cell versus a peptide that can only translocate across a dead or dying cell. Despite these differences, we have seen localization patterns in E. coli spheroplasts for many AMPs that are consistent with known mechanisms of action15, and our results for BF2 P11A with B. megaterium protoplasts presented here seem consistent with other observations for that peptide13,27,28. The membrane compositions of spheroplasts and protoplasts are also certainly more physiologically relevant than the typical mixtures used in other model systems, such as lipid vesicles. In summary, given the enhanced resolution and ease of imaging afforded by spheroplasts and protoplasts, we believe they are useful models for visualizing membrane active agents, such as AMPs, in a range of bacterial strains.
The authors have nothing to disclose.
Research was supported by National Institute of Allergy and Infectious Diseases (NIH-NIAID) award R15AI079685.
Trizma hydrocloride (Tris HCl) | Sigma | T3253 | |
Trizma base (Tris OH) | Sigma | T1503 | |
Magnesium chloride | Sigma | M8266 | |
Sucrose | Sigma | S7903 | |
Lysozyme | Sigma | L6876 | |
Deoxyribonuclease I | Sigma | D4527 | |
Ethylenediaminetetraacetic acid | Sigma | 106361 | Used Sigma 106361 in original protocol development; 106361 discontinued with ED2SS as replacement |
Cephalexin hydrate | Sigma | C4895 | |
Ampicillin | Fisher Scientific | BP1760 | |
BBL Trypticase soy broth | Fisher Scientific | B11768 | |
BF2 P11A FITC | NeoScientific | Custom ordered | |
di-8-ANEPPS | Biotium | 61012 | |
DMSO | Sigma | 34869 | Used Sigma D8779 in original protocol development; D8779 discontinued with 34869 as replacement |
Maleic acid | Sigma | M0375 | |
Acrodisc 25 mm Syringe Filter w/ 0.2 μm HT Tuffryn Membrane | Pall Corporation | 4192 | |
Laser scanning confocal microscope | Leica Microsystems | TCS SP5 II | For image acquisition |
Leica Application Suite, Advanced Fluorescence | Leica Microsystems | For image processing |