Gram's Stain Does Not Cross the Bacterial Cytoplasmic Membrane

Apr 16, 2015 - For well over a century, Hans Christian Gram's famous staining protocol has been the standard go-to diagnostic for characterizing unkno...
0 downloads 15 Views 602KB Size
Subscriber access provided by EMORY UNIV

Article

Gram’s stain does not cross the bacterial cytoplasmic membrane Michael J. Wilhelm, Joel B. Sheffield, Mohammad Sharifian Gh., Yajing Wu, Christian Spahr, Grazia Gonella, Bolei Xu, and Hai-Lung Dai ACS Chem. Biol., Just Accepted Manuscript • DOI: 10.1021/acschembio.5b00042 • Publication Date (Web): 16 Apr 2015 Downloaded from http://pubs.acs.org on April 18, 2015

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Chemical Biology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

Gram’s stain does not cross the bacterial cytoplasmic membrane ,†







†,§

Michael J. Wilhelm* , Joel B. Sheffield , Mohammad Sharifian Gh. , Yajing Wu , Christian Spahr , †,|| † ,† Grazia Gonella , Bolei Xu , and Hai-Lung Dai* † ‡

th

Department of Chemistry, Temple University, 1901 N. 13 Street, Philadelphia, PA 19122, USA. th Department of Biology, Temple University, 1900 N. 12 Street, Philadelphia, PA 19122, USA.

ABSTRACT For well over a century, Hans Christian Gram’s famous staining protocol has been the standard go-to diagnostic for characterizing unknown bacteria. Despite continuous and ubiquitous use, we now demonstrate that the current understanding of the molecular mechanism for this differential stain is largely incorrect. Using the fully complementary time-resolved methods: second-harmonic light-scattering and bright-field transmission microscopy, we present a realtime and membrane specific quantitative characterization of the bacterial uptake of crystal-violet (CV), the dye used in Gram’s protocol. Our observations contradict the currently accepted mechanism which depicts that, for both Gram-negative and Gram-positive bacteria, CV readily traverses the peptidoglycan mesh (PM) and cytoplasmic membrane (CM) before equilibrating within the cytosol. We find that, not only is CV unable to traverse the CM, but on the time-scale of the Gram-stain procedure, CV is kinetically trapped within the PM. Our results indicate that CV, rather than dyes which rapidly traverse the PM, is uniquely suited as the Gram-stain. KEYWORDS Gram-stain mechanism; membrane transport; molecular uptake; second-harmonic generation; bacteria

ACS Paragon Plus Environment

1

ACS Chemical Biology

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 21

INTRODUCTION The Gram-stain1,2 is a classic biological protocol that is still actively used to differentiate bacteria into two possible classifications3–11: Gram-positive (Gram+) cells, in which the stain is retained, and Gram-negative (Gram-) cells, in which the stain is lost. The bacterial response to the Gramstain method, however, is not uniform, as some bacteria exhibit so-called Gram-variability in which cells of seemingly identical composition yield a mixed stain response12,13. Nevertheless, concerted experimental efforts have deduced that Gram+ and Gram- bacteria differ principally in their cellular ultrastructure12. Specifically, as depicted in Figure 1, Gram- cells are composed of a pair of distinct lipoprotein membranes: a lipopolysaccharide (LPS) coated outer membrane (OM) and an inner cytoplasmic membrane (CM), which are separated by a peptidoglycan mesh (PM) that is bound to the OM through a series of peptidoglycan-associated lipoproteins (PaL)12,14. Conversely, Gram+ cells are comparatively simpler and posses only a single lipoprotein membrane (i.e., the CM), though their PM is typically ca. 10-20+ times thicker than that found in Gram- cells12. The modern Gram-stain protocol13 consists of a series of time-sensitive steps: 1.) Staining heat-fixed cells with a solution of crystal violet (CV), where the positively charged CV molecules passively diffuse into the cell and electrostatically bind to available anionic surfaces; 2.) Introduction of a mordant (typically a solution of iodine and potassium iodide) to react with cationic CV, yielding a CV-mordant precipitate; 3.) An alcohol wash to remove stain from Gramcells (i.e., decolorization); 4.) Counter staining of the Gram- cells (typically using the red dye, safranin O); and finally 5.) Differentiating stained cells using optical microscopy. Numerous variations to this protocol, specifically using either a modified CV or mordant have been introduced to permit alternative methods of characterization. For example, Davies et al. introduced a mordant which allowed characterization by scanning transmission electron microscopy (STEM) and energy dispersive x-ray analysis (EDS)15,16. More recently, Budin et al. introduced a chemically modified CV containing a magnetic fluorescent nanoparticle, allowing characterization with either fluorescence or magnetic resonance17. The currently accepted molecular mechanism of the Gram-stain (Figure S1, see Supporting Information, SI, for details) stems predominantly from Davies and coworker’s STEM-EDS experiments12,15,16. Specifically, it has been proposed that both CV and the mordant freely cross the three main cellular barriers (i.e., OM, PM, and CM), resulting in accumulation of the CV precipitate within the periplasm and the cytosol. Next, as alcohol denatures proteins and dissolves lipid membranes18, the subsequent decolorization step therefore attacks the structural integrity of both the OM and CM. Of significance, as the OM is covalently bound to the PM, the

ACS Paragon Plus Environment

2

Page 3 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

PM (of Gram- species) is significantly disrupted when the OM is torn away. Conversely, the thick PM of Gram+ species is only slightly perturbed, resulting in local bulges and deformations, but remains largely preserved. Subsequently, the CV precipitate stain is readily washed out of the perforated PM of the Gram- species, but retained by the intact PM of the Gram+ species. We now demonstrate that, in contrast to the conventional understanding that CV crosses all cellular barriers, the Gram-stain method actually works because CV slowly diffuses across the PM and does not penetrate the cytoplasmic membrane. Transport of molecules across living cell membranes can be observed with real time resolution through time-resolved second-harmonic light scattering (SHS)19,20 and bright-field optical transmission microscopy (TM)20. Specifically, SHS is a surface sensitive probe based on the phenomenon second-harmonic generation (SHG) in which a portion of an incoming light is converted to light of twice the original frequency by interacting with a medium21,22. Molecules which exhibit non-zero hyperpolarizability, generally characterized by molecular structures lacking inversion symmetry, are capable of producing SHG. In freely rotating media (e.g., liquids), such molecules are randomly oriented and therefore yield no detectable SHG, as SH fields from isotropically oriented molecules cancel with one another. However, if such molecules adsorb onto a surface or interface (and are aligned with each other due to local surface interactions), the emitted SH fields from the aligned molecules interfere constructively to produce a coherent SH signal21–25. This principle has been applied for detecting molecules adsorbed on surfaces of colloidal particles (e.g., cells, nanoparticles), in which the emitted light is scattered in angular patterns characteristic of particle size22,23,26,27. For membrane systems (e.g., cells and liposomes), molecules adsorbed on the opposing leaflets of a bilayer have opposite orientations. Subsequently, SH fields from molecules adsorbed on the opposing surfaces (separated by a distance much shorter than the optical wavelength) would be out of phase and cancel with one another. The SHS response of an SHactive species traversing a membrane is therefore characterized by an initial rise, corresponding to adsorption onto the outer surface, followed by a decay, indicating the molecules have crossed the membrane and adsorbed onto the opposite interior surface. The resulting kinetic trace yields the characteristic SHS transport peak, the width of which is inversely proportional to the crossing rate. This behavior has previously been verified by observations made on liposomes28–38, and was the basis for our study of molecular uptake in Escherichia coli (E.coli)19,20 and murine erythroleukemia cells39. Similarly, real-time bright-field TM can be used to quantitatively monitor molecular uptake by measuring the attenuation of transmitted light via a simple Beer’s law relationship20. As the

ACS Paragon Plus Environment

3

ACS Chemical Biology

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 21

concentration of a molecule of interest within a cell increases, less light is transmitted through the cell, resulting in a darker image. By monitoring the time-dependent absorption from a region of interest (ROI) within an imaged cell, a kinetic trace of molecular uptake can be obtained, where accumulation and saturation/equilibration of a molecule of interest is revealed in the timeresolved signal. Of significance, as the TM signal is dependent solely on the concentration (i.e., rather than the orientation), TM can be used as a check for a possible false-negative SHS result. For example, the absence of a SHS transport peak can either imply a lack of transport or a non-anisotropic surface ordering. Using TM to measure the equilibration time of the molecule within the cell provides a direct check of the validity of the measured SHS results. In this report, we present a characterization of the real-time molecular uptake of CV into living E. coli using time-resolved SHS and TM as a test of the validity of the currently accepted Gram-stain mechanism inferred from the results of Davies and colleagues15,16. Significantly, we present direct evidence proving that CV is incapable of crossing the CM. Furthermore, on the time-scale of the Gram-stain protocol, our results indicate that CV barely diffuses beyond the thin Gram- PM. As a control test of the CV results, we also examine the transport behavior of a structurally similar TPM dye, malachite green (MG), which is known to cross the bacterial OM, PM, and CM. Finally, simultaneous quantitative analysis of the SHS and TM molecular uptake kinetic responses allows a direct measure of the associated CV transport rates in E. coli. RESULTS and DISCUSSION Both MG and CV cross the OM, but only MG crosses the CM. We have previously characterized the molecular uptake of MG into living bacteria using the complementary methods SHS19,20 and TM20 (Figure 2). Real-time TM shows that MG saturates the bacteria over the course of roughly 8 minutes, in which individual TM images are shown to grow progressively darker as transmission of broadband visible light is attenuated by increasing local concentrations of MG within the cell (Figure 2a). ROI average absorption from each of the TM images, plotted as a function of time, reveals a monotone increasing response followed by a late-time plateau (Figure 2b, blue curve), suggesting equilibrium of MG at accessible cellular surfaces and within sampled internal domains. Similarly, membrane-specific SHS reveals two distinct transport events (Figure 2c, green curve), where transport is characterized as a sequential rise and decay of the SHS signal. Recall that Gram- bacteria, such as E. coli, are composed of dual membranes (i.e., OM and CM). Subsequently, the observation of two discrete transport events (Figure 2c, green curve), coupled with the corroborating TM deduced ca. 8

ACS Paragon Plus Environment

4

Page 5 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

minute saturation delay (Figure 2b, blue curve), suggests MG traverses both the OM and CM and equilibrates within the cytosol. In contrast to MG, uptake of the structurally similar Gram’s stain dye, CV, displays remarkable differences. Similar to MG, time-resolved TM shows that CV also saturates the bacteria, but in a mere 2 minutes (i.e., four times faster than MG, Figure 2a and 2b). Further, SHS from CV reveals only a single transport peak (Figure 2c, purple curve), identical to the fast transport peak observed for MG (Figure 2c, inset), and is suggestive of rapid transport across the water filled porin channels embedded in the OM. However, following fast transport across the OM, CV exhibits significantly slower PM diffusion-limited adsorption onto the outer leaflet of the CM, as indicated by the comparatively slow secondary rise of the SHS signal. Further, following adsorption onto the CM, the SHS signal simply exhibits a late time plateau (Figure 2c). Recall that attenuation of the SHS signal is indicative of transport across the membrane, as adsorption onto the opposite leaflet of the membrane results in the cancellation of the emitted SH fields. Subsequently, the SHS plateau indicates that CV does not cross the CM. This interpretation is confirmed by the time-resolved TM results, which suggests that CV equilibrates within the bacteria roughly four times faster than MG. This is reasonable if uptake of CV is limited to sampling only the periplasmic space between the OM and CM, but not the cytosol. Subsequently, as opposed to MG, CV is observed to slowly diffuse through the PM and is unable to traverse the CM. Heat fixation does not alter CM permeability. It is important to consider the relative role that heat fixation plays in the Gram-stain protocol. Of significance, the addition of a sufficient quantity of heat to a living cellular sample could eventually lead to denaturation of proteins, resulting in perforated membranes as proteins fall out of place, and hence an artificial enhancement of membrane permeability. We note that the SHS experiments employed here used a continuously flowing liquid jet of the bacterial suspension (see Methods). Subsequently, to ensure identical samples for both SHS and TM experiments, neither of the samples were heat fixed. In order to establish the validity of comparing our sample preparation against a traditional heat fixed sample, we characterized the relative variation of membrane permeability (i.e., following heat fixation) using a propidium iodide (PI) fluorescence test40. Specifically, PI is known to exhibit enhanced fluorescence following intercalation with DNA. As bacterial DNA is found exclusively within the cytosol, and as PI is unable to traverse the CM of viable cells, fluorescence is only observed for cells with compromised membranes (e.g., dead cells). Subsequently, if heat fixation does enhance the permeability of the CM (i.e., due to denatured

ACS Paragon Plus Environment

5

ACS Chemical Biology

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 21

membrane proteins), PI stained samples would exhibit a dramatic increase in the number of fluorescing cells. Figure 3 depicts typical bright-field and fluorescence images for heat fixed and unheated samples. Of significance, both sample preparations exhibit roughly 25% fluorescing cells. This suggests that heat fixation does not notably influence the permeability of the bacterial CM. Therefore, it is reasonable to directly compare results from our unheated samples against traditional heat fixed Gram-stain preparations. PM diffusion is the rate limiting step in bacterial uptake of CV. The SHS and TM results describe distinct aspects of the same system response, e.g. uptake of CV into E. coli via adsorption and transport across the cellular barriers, and can therefore be quantitatively described by a common kinetic model that describes the process. This model can thus be used as a basis for non-linear least squares fitting analysis (see SI for details) of both SHS and TM observations in real time. Briefly, accounting for: 1.) Adsorption and desorption from the outer leaflet of the OM:  +

 ⇌  

+  , 

 



(1)

where  is the extracellular CV concentration,  

and  

are empty and filled surface sites on the outer leaflet of the OM, and  and  are rate constants for CV adsorption /

desorption at the outer leaflet of the OM; 2.) Porin-assisted transport across the OM:   ,  ⇌    

(2)

where  is the CV concentration in the volume spanning the OM and PM, and   and    are the forward and reverse crossing rates for the OM porin channels;

3.) Adsorption and desorption from the inner leaflet of the OM:   + 

⇌ 

+  , 

(3)

where 

and 

are empty and filled surface sites on the inner leaflet of the OM, and

 and  are rate constants for CV adsorption / desorption at the inner leaflet of the OM; 4.) Diffusion across the PM:   ⇌  ,  



ACS Paragon Plus Environment

(4)

6

Page 7 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

  where  is the CV concentration in the volume spanning the PM and CM, and  and 

are the forward and reverse crossing rates for the PM; and 5.) Adsorption onto the outer leaflet of the CM:      +  

⇌  

+  , 

(5)

  where  

and  

are empty and filled surface sites on the outer leaflet of the CM, and

 and  are the same rate constants for CV adsorption / desorption at the inner leaflet of the OM; yields a series of coupled differential equations whose solutions describe the time dependent evolution of the concentration of CV on available surfaces (i.e.,  



, and     

) and within accessible cellular compartments (i.e.,  ,  , and  ). As shown

previously, time-resolved SHS and TM provide complementary measures of uptake, hence the resulting signals can be simultaneously analyzed using the series of coupled equations (see SI for details). Figure 4 depicts the simultaneous fits for the TM and SHS responses of CV interacting with E.coli, along with a schematic (Figure 4b) comparing uptake rates for MG and CV (deduced in Figure 4a). The deduced transport rates for MG and CV crossing the OM, PM, and CM are tabulated in Table 1. As shown in the inset of Figure 2c and Table 1, both CV and MG traverse the OM with effectively identical rates (within error), likely through the OM porin channels. Conversely, CV is shown to traverse the PM roughly 50 times slower than MG! Updated molecular mechanism of the Gram-Stain. In this report, we used the complementary methods of SHS and TM to quantitatively characterize the real-time bacterial uptake of CV as a direct test of the currently accepted molecular mechanism of the Gram-stain protocol. In stark contrast to the currently accepted description, our results clearly demonstrate that CV is unable to cross the bacterial CM. Furthermore, we demonstrate that, on the timescale of the Gram-stain protocol, CV is barely able to diffuse beyond the comparatively thin PM of Gram- bacteria. Given the substantially thicker PM of Gram+ cells, CV is invariably kinetically trapped within the PM. In light of the revelation that CV does not cross the CM, a re-examination of the results of Davies et al.15,16 is in order. It must be pointed out that the signal detected in their STEM-EDS experiments was sensitive solely to the Pt in the mordant, and not the CV. Specifically, it was simply assumed that the detected Pt signal originated from mordant that had reacted with CV and precipitated out of solution. However, there is no reason to preclude the possibility that the

ACS Paragon Plus Environment

7

ACS Chemical Biology

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 21

mordant alone was capable of traversing the CM and therefore producing the observed signal. In light of the fact that control experiments testing the uptake of the mordant in the absence of CV were not performed, this possibility cannot be ruled out. Nevertheless, the observations regarding perforation of the OM, PM, and CM of Gram- species, and minor perturbations to the Gram+ PM remain valid and useful for garnering insights to the Gram-stain mechanism. Specifically, as corroborated by our current SHS results, CV diffuses surprisingly slow through the PM (e.g., fifty times slower than MG). Subsequently, on the time-scale of the Gram-stain protocol, CV is trapped within the PM. Therefore, given that the Gram- PM is violently perforated in the decolorization step, it makes perfect sense that the CV-mordant precipitate is lost. Conversely, as the Gram+ PM is only slightly perturbed, it is likewise fitting that the CV-mordant precipitate is retained. An updated molecular mechanism of the Gram-stain protocol is presented in Figure 5. It should be noted that our model does not negate the utility of the protocol, nor invalidate prior applications of the stain. Rather, our model provides new insight into why the protocol actually works, and therefore how it can be improved. As distinct from the currently accepted model (Figure S1), our results demonstrate that the differential behavior of the Gram-stain protocol originates from the diffusion limiting interaction of CV with the PM. Of significance, this suggests that CV is uniquely suited for this purpose, and that structurally similar dyes, such as MG (i.e., which rapidly traverse the PM), are ineffective substitutes. Subsequently, the development of modified dyes/mordants to allow differential characterization via alternative methodologies15–17 must conserve the slow PM diffusion rate. Furthermore, our results hint that the Gram-variable response likely originates from variations of the PM. Finally, given the structural similarity of the two TPM dyes, it is worth examining the molecular properties which give rise to the remarkably different membrane transport responses. In addition to their similar size and mass, both CV and MG possess three large phenyl rings, and should presumably be capable of existing within the hydrophobic interior of a phospholipid membrane. Further, both exhibit a singular cationic charge, thought to be localized on the central carbon. However, as depicted in the inset of Figure 2b, the CV cation is of slightly higher symmetry than MG (i.e., C3V vs. C2V). Correspondingly, whereas the two main resonance structures of cationic MG yield a permanent non-zero dipole moment, the three identical structures for CV do not. Subsequently, our results highlight the possible necessity of a permanent dipole for the passive membrane transport of molecular ions. Nevertheless, further experiments are necessary before such a generalization can be established.

ACS Paragon Plus Environment

8

Page 9 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

In summary, we have used the real-time complementary methods of SHS and TM to quantitatively characterize the membrane specific uptake of the Gram-stain. In contrast to the previously accepted molecular mechanism, we definitively demonstrate that CV is unable to cross the bacterial CM. Moreover, on the time-scale of the Gram-stain protocol, we show that CV is effectively trapped within the PM. Subsequently, the differentiability of the Gram-stain resides predominantly in the stability of the PM: i.e., if the PM is destroyed, the stain is lost; if the PM is preserved, the stain is retained. METHODS Sample preparation. Discrete colonies of E. coli (mc4100 strain, ATCC 35695) were grown on Luria Broth agar (Sigma-Aldrich) plates. Experimental samples, prepared from single colonies, were cultured (37oC, 150 RPM shaking, ca. 18 hours) in 100 mL of Terrific Broth (TB, Sigma Aldrich) to late-log/early-stationary phase. Samples were lightly pelletized by centrifugation (ca. 1500xg for 10 minutes), and washed in 1xPBS, three times to remove waste and residual TB. Washed pellets were re-suspended in 1xPBS to a stock cell density of ca. 2x108 cells ml-1. Concentrated stock solutions of MG and CV were prepared by dissolution of the oxalate and chloride salt, respectively, used as obtained from the supplier (Sigma-Aldrich). Final sample concentrations were maintained at 10 µM (MG) and 50 µM (CV).

Second-harmonic generation scattering. The 800 nm output from a Ti:Sapphire laser (Coherent, Micra V, oscillator only, 50 fs pulse duration, 4 nJ pulse energy, 76 MHz repetition rate, 0.4 W average power) was used as a fundamental light source. To minimize laser absorption losses, multiple scattering effects, etc., SHS was measured while the sample circulated in a liquid flow system. A continuous liquid jet was formed by pumping the sample through a circular stainless-steel nozzle (1/1600 inner diameter). Nalgene tubing (Nalge Nunc, Inc.) was used both to connect a 40 mL triple-neck round-bottom flask sample reservoir with the inlet of a motorized liquid pump (Micropump, Inc.), as well as to recollect the sample back into the reservoir. To ensure collection of SH signal solely from the sample, a long-pass filter (Schott, RG695) was placed in front of the focusing lens immediately before the sample jet. Further, as both fundamental (800 nm) and SH light (400 nm) is scattered from the sample, a BG39 band-pass filter and monochromator (1 mm entrance and exit slits, 400±1 nm bandwidth) were used to selectively collect the SH signal. The signal was then detected by a photomultiplier (Hamamatsu, R585), pre-amplified (Stanford Research Systems, SR 440), and processed through a correlated photon counting system (Stanford Research Systems, SRS SR400).

ACS Paragon Plus Environment

9

ACS Chemical Biology

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 21

To separate the SHG response from the hyper-Rayleigh scattering (HRS) of the bulk dye solution, SHS was first measured from the dye solution in the absence of cells. An aliquot of the concentrated cellular stock solution was then added to the reservoir, allowing SHG to be measured as a perturbation to the baseline HRS signal. To ensure rapid mixing of the dye/cell suspension, the contents of the sample reservoir were continuously stirred. Bright-field transmission microscopy imaging. All images were collected in brightfield transmission mode using a Leica DMRXE microscope with a 100x PlanApo objective lense coupled to a digital image capture system (Tucsen model TC-3), software controlled with TSView (OnFocus Labs, ver. 7). Images, covering a field of view (FOV) of ca. 60x50 µm2, were collected in 2 second increments, typically over a total duration of 10 minutes, with a fixed resolution of 2048x1536 pixels (ca. 31 nm/pixel), and saved using the tagged image file format (TIFF). Sample slides were prepared as follows: Transparent masking tape was used to construct a ‘U’ shaped sample reservoir with approximate physical dimensions of 12 x 25 x 0.1 mm3. The reservoir was loaded with 0.015 mL of a solution of bacteria suspended in 1xPBS (ca. 6x106 cells mL-1), and covered with a glass coverslip. Immobilization of cells was achieved via one of three methods: 1.) native electrostatic interactions between the anionic cell surface and the cationic glass slide, 2.) enhanced electrostatic interactions between the cell and a poly-llysine (Sigma) treated slide, and 3.) gentle heat fixation of cells dried on the slide, followed by re-hydration with an equal volume of distilled water. After mounting the slide onto the microscope, image acquisition was started. After ca. 10 seconds of acquisition, a 0.015 mL aliquot of dye (with either CV or MG) was added at the edge of the coverslip, which then diffused beneath and began staining the cells. Image analysis, including image inverse and quantification of the mean and standard deviation of region-of-interest (ROI) signal intensity, was performed in ImageJ (National Institutes of Health, 1.43u). Average time-resolved TM signal was determined as the mean of a series of ROI average inverted intensities, sampled over ten cells, spanning three separate experiments. Fluorescence microscopy of PI stained cells. Sample slides were prepared as described above using either heat fixation or poly-l-lysine prepared slides to immobilize bacterial cells. Cells were stained with 0.015 mL of 10 µM PI (Sigma) for 10 minutes. For each bright-field transmission image acquired, a corresponding fluorescence image (λexcitation = 525-550 nm, λdetection > 590 nm) was collected immediately afterwards. Relative percent fluorescing cells were determined by direct count in ImageJ (National Institutes of Health, 1.43u).

ACS Paragon Plus Environment

10

Page 11 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

ASSOCIATED CONTENT Additional figures and a detailed kinetic model of molecular uptake: This material is available free of charge via the Internet at http://pubs.acs.org. AUTHOR INFORMATION Corresponding Authors * M.J.W., [email protected]; H.-L.D., [email protected] Present address §

C.S., Department of Biology, Eidgenössische Technische Hochschule (ETH) Zürich, Zürich.

Switzerland. ||

G.G., Max Planck Institute for Polymer Research, Ackermannweg 10, 55128 Mainz, Germany.

Notes The authors declare no competing financial interest. AUTHOR CONTRIBUTIONS M.J.W., J.B.S., and H.-L.D. conceived and designed the experiments; M.J.W., J.B.S., Y.W., C.S. and M.S.G. performed the time-resolved TM experiments; M.J.W., Y.W., C.S., M.S.G., and B.X. performed the time-resolved SHS experiments; M.J.W. and M.S.G. performed the PI fluorescence experiments; M.J.W., Y.W., C.S., M.S.G., and G.G. analyzed the data; M.J.W. and H.-L.D. interpreted the results; and M.J.W. and H.-L.D. wrote the paper with input from the other authors. ACKNOWLEDGEMENTS We are grateful to S. Paudyal and A. Nicholson of Temple University for assistance with cell culturing techniques. This work was supported by the National Science Foundation (Grant No. CHE-1058883). REFERENCES (1) Friedlander, C. (1883) Die mikrokokken der pneumonie. Fortschr. Med. 1, 715–733. (2) Gram, H. C. (1884) Über die isolierte färbung der schizomyceten in schnitt- und trockenpräparaten. Fortschr. Med. 2, 185–189. (3) Mathews, C. J., Weston, V. C., Jones, A., Field, M., and Coakley, G. (2010) Bacterial septic arthritis in adults. Lancet 375, 846–855.

ACS Paragon Plus Environment

11

ACS Chemical Biology

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 21

(4) Herrera, A. F., Soriano, G., Bellizzi, A. M., Hornick, J. L., Ho, V. T., Ballen, K. K., Baden, L. R., Cutler, C. S., Antin, J. H., Soiffer, R. J., and Marty, F. M. (2011) Cord colitis syndrome in cord-blood stem-cell transplantation. N. Engl. J. Med. 365, 815–824. (5) Ege, M. J., Mayer, M., Normand, A.-C., Genuneit, J., Cookson, W. O. C. M., Braun-Fahrlander, C., Heederik, D., Piarroux, R., and von Mutius, E. (2011) Exposure to environmental microorganisms and childhood asthma. N. Engl. J. Med. 364, 701–709. (6) Panizzi, P., Nahrendorf, M., Figueiredo, J.-L., Panizzi, J., Marinelli, B., Iwamoto, Y., Keliher, E., Maddur, A. A., Waterman, P., Kroh, H. K., Leuschner, F., Aikawa, E., Swirski, F. K., Pittet, M. J., Hackeng, T. M., Fuentes-Prior, P., Schneewind, O., Bock, P. E., and Weissleder, R. (2011) In vivo detection of Staphylococcus aureus endocarditis by targeting pathogen-specific prothrombin activation. Nat. Med. 17, 1142–1146. (7) Brouwer, M. C., Thwaites, G. E., Tunkel, A. R., and van de Beek, D. (2012) Dilemmas in the diagnosis of acute community-acquired bacterial meningitis. Lancet 380, 1684–1692. (8) Broz, P., Ruby, T., Belhocine, K., Bouley, D. M., Kayagaki, N., Dixit, V. M., and Monack, D. M. (2012) Caspase-11 increases susceptibility to Salmonella infection in the absence of caspase-1. Nature 490, 288–291. (9) Chiang, N., Fredman, G., Bäckhed, F., Oh, S. F., Vickery, T., Schmidt, B. a, and Serhan, C. N. (2012) Infection regulates pro-resolving mediators that lower antibiotic requirements. Nature 484, 524–528. (10) Yipp, B. G., Petri, B., Salina, D., Jenne, C. N., Scott, B. N. V, Zbytnuik, L. D., Pittman, K., Asaduzzaman, M., Wu, K., Meijndert, H. C., Malawista, S. E., de Boisfleury Chevance, A., Zhang, K., Conly, J., and Kubes, P. (2012) Infection-induced NETosis is a dynamic process involving neutrophil multitasking in vivo. Nat. Med. 18, 1386–1393. (11) Conlon, B. P., Nakayasu, E. S., Fleck, L. E., LaFleur, M. D., Isabella, V. M., Coleman, K., Leonard, S. N., Smith, R. D., Adkins, J. N., and Lewis, K. (2013) Activated ClpP kills persisters and eradicates a chronic biofilm infection. Nature 503, 365–370. (12) Popescu, A., and Doyle, R. J. (1996) The Gram stain after more than a century. Biotech. Histochem. 71, 145–151. (13) Moyes, R. (2009) Differential staining of bacteria: Gram stain. Curr. Protoc. Microbiol. 15, A.3C.1– A.3C.8. (14) Godlewska, R., Wiśniewska, K., Pietras, Z., and Jagusztyn-Krynicka, E. K. (2009) Peptidoglycanassociated lipoprotein (Pal) of Gram-negative bacteria: function, structure, role in pathogenesis and potential application in immunoprophylaxis. FEMS Microbiol. Lett. 298, 1–11. (15) Davies, J. A., and Anderson, G. K. (1983) Chemical mechanism of the Gram stain and synthesis of a new electron-opaque marker for electron microscopy which replaces the iodine mordant of the stain. J. Bacteriol. 156, 837–845. (16) Beveridge, T. J., and Davies, J. A. (1983) Cellular responses of Bacillus subtilis and Escherichia coli to the Gram stain. J. Bacteriol. 156, 846–858. (17) Budin, G., Chung, H. J., Lee, H., and Weissleder, R. (2012) A magnetic Gram stain for bacterial detection. Angew. Chem. Int. Ed. 51, 7752–7755.

ACS Paragon Plus Environment

12

Page 13 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

(18) Nozaki, Y., and Tanford, C. (1971) The solubility of amino acids and two glycine peptides in aqueous ethanol and dioxane solutions: Establishment of a hydrophobicity scale. J. Biol. Chem. 246, 2211–2217. (19) Zeng, J., Eckenrode, H. M., Dounce, S. M., and Dai, H.-L. (2013) Time-resolved molecular transport across living cell membranes. Biophys. J. 104, 139–145. (20) Wilhelm, M. J., Sheffield, J. B., Gonella, G., Wu, Y., Spahr, C., Zeng, J., Xu, B., and Dai, H.-L. (2014) Real-time molecular uptake and membrane-specific transport in living cells by optical microscopy and nonlinear light scattering. Chem. Phys. Lett. 605-606, 158–163. (21) Shen, Y. R. (1989) Surface properties probed by second-harmonic and sum-frequency generation. Nature 337, 519–525. (22) Roke, S., and Gonella, G. (2012) Nonlinear light scattering and spectroscopy of particles and droplets in liquids. Annu. Rev. Phys. Chem. 63, 353–378. (23) Gonella, G., and Dai, H.-L. (2014) Second harmonic light scattering from the surface of colloidal objects: Theory and applications. Langmuir 30, 2588–2599. (24) Wang, H., Yan, E. C. Y., Borguet, E., and Eisenthal, K. B. (1996) Second harmonic generation from the surface of centrosymmetric particles in bulk solution. Chem. Phys. Lett. 259, 15–20. (25) Wang, H.-F., Troxler, T., Yeh, A.-G., and Dai, H.-L. (2000) In Situ, nonlinear optical probe of surfactant adsorption on the surface of microparticles in colloids. Langmuir 16, 2475–2481. (26) Gonella, G., and Dai, H.-L. (2011) Determination of adsorption geometry on spherical particles from nonlinear Mie theory analysis of surface second harmonic generation. Phys. Rev. B 84, 121402. (27) Eisenthal, K. B. (2006) Second harmonic spectroscopy of aqueous nano- and microparticle interfaces. Chem. Rev. 106, 1462–1477. (28) Kim, J. H., and Kim, M. W. (2008) In-situ observation of the inside-to-outside molecular transport of a liposome. J. Phys. Chem. B 112, 15673–15677. (29) Liu, J., Subir, M., Nguyen, K., and Eisenthal, K. B. (2008) Second harmonic studies of ions crossing liposome membranes in real time. J. Phys. Chem. B 112, 15263–15266. (30) Liu, J., Shang, X., Pompano, R., and Eisenthal, K. B. (2005) Antibiotic assisted molecular ion transport across a membrane in real time. Faraday Discuss. 129, 291–299. (31) Liu, Y., Yan, E. C., and Eisenthal, K. B. (2001) Effects of bilayer surface charge density on molecular adsorption and transport across liposome bilayers. Biophys. J 80, 1004–1012. (32) Saini, R. K., Dube, A., Gupta, P. K., and Das, K. (2012) Diffusion of chlorin-p6 across phosphatidyl choline liposome bilayer probed by second harmonic generation. J. Phys. Chem. B 116, 4199–4205. (33) Shang, X., Liu, Y., Yan, E., and Eisenthal, K. B. (2001) Effects of counterions on molecular transport across liposome bilayer: Probed by second harmonic generation. J. Phys. Chem. B 105, 12816–12822. (34) Srivastava, A., and Eisenthal, K. B. (1998) Kinetics of molecular transport across a liposome bilayer. Chem. Phys. Lett. 292, 345–351.

ACS Paragon Plus Environment

13

ACS Chemical Biology

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 21

(35) Varshney, G. K., Saini, R. K., Gupta, P. K., and Das, K. (2013) Effect of curcumin on the diffusion kinetics of a hemicyanine dye, LDS-698, across a lipid bilayer probed by second harmonic spectroscopy. Langmuir 29, 2912–2918. (36) Yamaguchi, A., Nakano, M., Nochi, K., Yamashita, T., Morita, K., and Teramae, N. (2006) Longitudinal diffusion behavior of hemicyanine dyes across phospholipid vesicle membranes as studied by second-harmonic generation and fluorescence spectroscopies. Anal. Bioanal. Chem. 386, 627–632. (37) Yan, E. C. Y., and Eisenthal, K. B. (2000) Effect of cholesterol on molecular transport of organic cations across liposome bilayers probed by second harmonic generation. Biophys. J. 79, 898–903. (38) Kim, J. H., and Kim, M. W. (2007) Temperature effect on the transport dynamics of a small molecule through a liposome bilayer. Eur. Phys. J. E 23, 313–317. (39) Zeng, J., Eckenrode, H. M., Dai, H.-L., and Wilhelm, M. J. (2015) Adsorption and transport of charged vs. neutral hydrophobic molecules at the membrane of murine erythroleukemia (MEL) cells. Colloids Surf., B 127, 122–129. (40) Yeh, C. J. G., Hsi, B., and Faulk, W. P. (1981) Propidium iodide as a nuclear marker in immunofluorescence. II. Use with cellular identification and viability studies. J. Immunol. Methods 43, 269–275.

ACS Paragon Plus Environment

14

Page 15 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

Table 1 Kinetic parameters for uptake of MG and CV by E. coli. Dye a MG CV

-2

  x10 s 4.3 ± 0.7 4.7 ± 0.5

-1

-2

-1

 x10 s 7.2 ± 0.4 0.14 ± 0.03

-4

 x10 s 2.2 ± 0.11 -

-1

Deduced values expressed as mean±SD, obtained from experiments with n=3 independent bacterial a 20 cultures. MG kinetic parameters deduced in Ref. .

ACS Paragon Plus Environment

15

ACS Chemical Biology

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

139x94mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 16 of 21

Page 17 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

67x28mm (300 x 300 DPI)

ACS Paragon Plus Environment

ACS Chemical Biology

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

67x54mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 18 of 21

Page 19 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

67x47mm (300 x 300 DPI)

ACS Paragon Plus Environment

ACS Chemical Biology

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

67x55mm (300 x 300 DPI)

ACS Paragon Plus Environment

Page 20 of 21

Page 21 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

139x43mm (300 x 300 DPI)

ACS Paragon Plus Environment