The Transformation Experiment of Frederick Griffith I: Its Narrowing and Potential for the Creation of Novel Microorganisms

Abstract

The construction of artificial microorganisms often relies on the transfer of genomes from donor to acceptor cells. This synthetic biology approach has been considerably fostered by the J. Craig Venter Institute but apparently depends on the use of microorganisms, which are very closely related. One reason for this limitation of the "creative potential" of "classical" transformation is the requirement for adequate "fitting" of newly synthesized polypeptide components, directed by the donor genome, to interacting counterparts encoded by the pre-existing acceptor genome. Transformation was introduced in 1928 by Frederick Griffith in the course of the demonstration of the instability of pneumococci and their conversion from rough, non-pathogenic into smooth, virulent variants. Subsequently, this method turned out to be critical for the identification of DNA as the sole matter of inheritance. Importantly, the initial experimental design (1.0) also considered the inheritance of both structural (e.g., plasma membranes) and cybernetic information (e.g., metabolite fluxes), which, in cooperation, determine topological and cellular heredity, as well as fusion and blending of bacterial cells. In contrast, subsequent experimental designs (1.X) were focused on the use of whole-cell homogenates and, thereafter, of soluble and water-clear fractions deprived of all information and macromolecules other than those directing protein synthesis, including outer-membrane vesicles, bacterial prions, lipopolysaccharides, lipoproteins, cytoskeletal elements, and complexes thereof. Identification of the reasons for this narrowing may be helpful in understanding the potential of transformation for the creation of novel microorganisms.

Keywords: cellular heredity; horizontal gene transfer; science and technology studies; scientific reductionism; synthetic biology; transforming principle.

Figure 1

Artificial construction of bacterial cells as a strategy of synthetic biology for the generation of novel organisms as proposed by J. Craig Venter and coworkers. A number of key technologies for the synthesis of genomes have been introduced by the J Craig Venter Institute (JCVI) during the past two decades, which have considerably facilitated the development of bacteria harbouring chemically synthesized genomes. First, the DNA molecules generated with the aid of an inefficient and slow process were limited to max. 32 kb. Subsequently invented methods significantly accelerated the synthesis of larger genomes with greatly improved efficacy and culminated in the first complete chemical synthesis of a bacterial genome (M. genitalium) in 2008 [2,15]. This success was based on the assembly of cassettes that had been built up from overlapping synthetic oligonucleotides. Cloning of the bacterial genomic DNA, which encodes a tetracycline resistance marker (for survival in the presence of this antibiotic in the culture broth during transformation of yeast), together with 3–5 kb vector sequences (red), which operate as centromeric plasmids in yeast, was performed to yield the total assembly of those overlapping bacterial sub-genomic DNA sequences. The production process of the high number of bacterial synthetic genomic DNA sequences required for genome transplantation was accomplished by culturing of the yeast cells with the parked synthetic bacterial genome. Following preparation of the synthetic bacterial DNA molecules from yeast at large scale, genome transplantation leads to their introduction into an appropriate bacterial acceptor cell. As a result, the transplanted synthetic genome now manages to direct the genotype and phenotype of the newly emerging cell. For genome transplantation, yeast cells harbouring the completely synthesized bacterial donor genome are encapsulated by small blocks of low-melting-point agarose with zymolyase and proteinase K immersed. These two enzymes provoke the conversion of the yeast cells into spheroplasts, with the bacterial donor genome left in the interior of the agarose blocks. This procedure protects the bacterial genome from shearing forces during the purification process due to its gentle release from the melting agarose. Upon mixing of the released bacterial donor DNA and the bacterial cells (M. capricolosum) in the presence of polyethylene glycol (to increase membrane fluidity), as well as CaCl₂ (to neutralize the charge of the donor DNA), the donor genome is prepared for uptake by the acceptor bacterial cells (albeit at a rather low rate). After growth and division of those acceptor cells (which are of diploid nature for a short period of time only), the addition of tetracycline to the culture results in selection for only bacterial cells that express the tetracycline resistance marker encoded by the synthetic donor genome exclusively [15].

Figure 2

Stability and reversion of pneumococcal serotypes using the Griffith transformation experimental design 1.0 [21]. Upper section: Serotype switching of smooth pneumococci. A sample of the sputum of a patient suffering from lobar pneumonia was injected into a mouse. The propagated bacteria turned out to belong to one of the common serotypes of smooth pneumococci, such as S (I). Another sample of this sputum was mixed with anti-S (I) antiserum to prevent infection of a second mouse by S (I) upon injection of the mixture. This mouse was killed in the course of an infection by pneumococci of a different serotype (S (IV)). This experimental design was repeated with another sample of sputum, which was mixed with both anti-S (I) and anti-S (IV) antisera, and, again, caused the death of the injected mouse due to the emergence of pneumococci of serotype S (III). This experimental setting was used by Griffith to demonstrate that many pneumonia patients exhibit pneumococci of two or more different serotypes. Lower section: The loss or regaining of virulence by pneumococci. S (X) of a certain serotype were grown in the presence of the homologous anti-S (X) antiserum raised against the capsular lipopolysaccharide (LPS) or special agar-containing broth. This led to the loss of the capsule from the former virulent S pneumococci and, concomitantly, the generation of non-virulent encapsulated R (X) pneumococci. They differed considerably in their ability to resynthesize capsular material. Some of them spontaneously regained their virulent character for the mice, whereas others remained more stably in the non-virulent state. The latter, more stable R (X) were tested for serotype switching, which relied on the injection (s.c.) of either 1 cc or 50–100 cc of a fully grown pneumococcal culture into a mouse. Injection of the low dose apparently did not lead to disease, whereas the high dose frequently caused fatal infections by pneumococci, which displayed the same serotype as the original R strain. Apparently, even the stable R (X) pneumococci managed to revert to their S (X) counterparts if injected at high concentrations under appropriate conditions.

Figure 3

Schematic representation of the original Griffith transformation experiment (design 1.1) in 1928 (A; see Ref. [21]) and its canonical interpretation (B). (A) Donor pneumococci of serotype S (I), corresponding to 50 cc of broth culture, which had been attenuated by incubation with anti-S antiserum (in the presence of chocolate broth) or killed by heat (60 °C, 2–3 h), were injected in parallel with a limited amount of viable R (II) acceptor pneumococci. From the few mice that had been killed after the injection, S (I) pneumococci were isolated and grown in culture to fully virulent S (I). (B) Donor pneumococci of the (S) serotype (a) release fragments of their genome (brown circles) into the incubation/culture medium due to (partial) lysis and loss of cellular integrity in the course of their attenuation by heat (b) and subsequent exposure to the acceptor pneumococci. (c) Subsequent researchers [22] subjected the partially lysed donor pneumococci to centrifugation, heat exposure, chloroform extraction and/or protease treatment, etc. (design 1.2–1.X). The released and (eventually partially) purified DNA fragments or plasmids/episomes, possibly intrinsically contained in the donor pneumococci, were then taken up by acceptor pneumococci of the rough serotype (pink DNA circles) (d) by molecular mechanisms (e), which were characterized in detail. Following recombination of relevant DNA fragments into the genome of the acceptor pneumococci (brown and pink DNA circles) and their replication (f), the genes putatively encoded by the DNA fragments become expressed (g), including those engaged in the synthesis pathway of the specific LPS of the smooth serotype (pink circles). Consequently, capsule materials of the smooth serotype are formed ((h); for a review, see Refs. [23,24,25]). This sequence of events ultimately leads to the conversion of rough to smooth pneumococci and explains the inheritance of a “minor” difference, such as non-virulence vs. virulence, in bacteria.

Figure 4

Alternative interpretations of the Griffith transformation experiment (1.1). (A) The production of extracellular donor vesicles from the plasma membranes (PM) of donor bacteria (a) in the course of specific exocytotic or budding mechanisms have been amply documented (for details, see text below; for a review, see Refs. [26,27,28,29]). PM vesicles that may be released in response to environmental factors, such as mechanical stress (b), specifically harbour the LPS determining the smooth phenotype of the donor pneumococci and resist enzymic attacks from both nuclease and protease (c). Upon transfer of the PM vesicles from the smooth donor (d) to rough acceptor pneumococci (e), they fuse with the PM of the latter (f). In contrast, the released vesicles are removed by centrifugation (100,000× g-pellet); solubilized by extraction with chloroform; degraded by nuclease, phospholipase or protease digestion; or destroyed by heat treatment (80 °C, one hour) (c). In the course of fusion of the vesicles donating the smooth LPS (for a review, see Refs. [24,25]) to the PM of rough acceptor pneumococci (d) and subsequent replication (g) by a molecular mechanism that may resemble the so-called PM memory [16], the acceptor pneumococci gradually switch their phenotype from rough to smooth until completion of transformation (h). Complex (e.g., morphological) features of pneumococci other than those mediating the smooth phenotype and not assayed by the transformation experiment (1.0–1.X) may be inherited by PM vesicles (i; for details, see below; for a review, see Refs. [30,31,32,33,34,35]). (B) Apparently, the R (II) bacteria manage to convert to virulent S (I) and initiate reformation of a capsule, starting from an intrinsic precursor moiety of the acceptor bacteria and completing its final structure in S (I)-typical fashion (blue arrows). Transformation is provoked with the aid of DNA, and/or (putatively) micelle-like lipoprotein/lipopolysaccharide (LP/LPS) complexes and/or PM vesicles. Both the complexes and vesicles could correspond to “signal”, “food” or “pabulum”, as offered by Griffith as an explanation. According to an alternative theoretical option, “blending”, “fusion” or “mixing” of attenuated S (I) and non-virulent R (II) bacteria may result in hybrid pneumococci displaying both (S) and (R) characteristics, with the virulent phenotype dominating (brown arrows).

Figure 5

Elimination of the need for passage through the mouse, as well as of the requirement of intact donor bacteria, for the demonstration of transformation (experimental design 1.2; see Refs. [70,71,72,73,74]). A small number of heat-killed pneumococci of serotype S (III), representing a total whole-cell homogenate of the donor bacteria, was incubated (48 h) with a small number of living acceptor pneumococci of serotype R (II) in the presence of mouse serum (to mimic the animal environment), bacteriological culture medium and a minute concentration of anti-R antiserum (to agglutinate the R pneumococci). This resulted in the emergence of pneumococci of serotype S (III) in some experiments, which was demonstrated by subsequent culturing and immunological analysis of the blood of infected mice. In contrast, the supernatant of a high-speed centrifugation of the total homogenate that had been prepared from the heat-killed S (III) pneumococci or the total homogenate generated by up to 23 cycles of freezing and thawing or preparations of the specific LPS purified from the total homogenate did not support the conversion of R (II) into S (III) pneumococci. PM vesicles or micelle-like LP/LPS-complexes are thought to be liberated from the pelleted, then resuspended S (III) donor pneumococci into the total (whole-cell) homogenates and to retain their putative transforming activity in the course of moderate heating (as is certainly the case for DNA and LPS) but may be eliminated from the supernatant of a high-speed centrifugation (PM vesicles), as well as by multiple cycles of freezing and thawing. Moreover, purified LPS turned out to be ineffective in eliciting transformation.

Figure 6

Demonstration of the operation of a typical homogenate/soluble fraction supporting transformation (experimental design 1.3; see Refs. [75,76]). Total (whole-cell) homogenates were prepared from pneumococci of the S (III) or S (I) serotype by using a limited number of cycles of freezing and thawing (rather than heating)—a considerably reduced number of cycles in comparison to that used by Dawson and Sia (see Figure 5). After heating of the total (whole-cell) homogenate and centrifugation for removal of crude particulate materials, the resulting supernatant was diluted and passed through a bacterial filter (made of porcelain and named a Berkefeld filter after its inventor). This procedure caused the removal of any bacteria while enabling the passage of any soluble materials (after adjustment of the pH of the homogenate to an alkaline level, thereby avoiding non-specific adsorption of its components by the porcelain material), thereby excluding the possibility that an occasional viable S pneumococcus in the homogenate could be responsible for the supposed positive transformation event. The filtered homogenate was concentrated about ten-fold and incubated with a small number of viable R (II) pneumococci in the presence of bacterial medium and anti-R antiserum.

Figure 7

Demonstration of the operation of a water-clear fraction supporting transformation (experimental design 1.4; see Refs. [75,76]). Total (whole-cell) homogenates were prepared from pneumococci of the S (III) or S (I) serotype by lysis with sodium deoxycholate and subsequent heating. After precipitation of most of the material released from the pneumococci with ethanol but leaving behind the bile salt, a pellet of a low-speed centrifugation was dissolved in salt solution. The material contained in this solution was precipitated and redissolved again without significant loss of transforming activity. Following treatment with powdered wood charcoal, the opalescent and weakly turbid solution was converted into a water-clear factor, putatively containing dispersed LP and LPS in detergent solution. Upon incubation with a small number of viable R (II) pneumococci in the presence of anti-R antiserum, this factor was capable of transforming them into S (III) or S (I) pneumococci, respectively. PM vesicles, as well as micelle-like LP/LPS complexes, were certainly eliminated from the water-clear factor in the course of the multi-stage preparation of the whole-cell homogenates, encompassing detergent solubilisation and subsequent purification of the factor. LPS and LP solubilized with deoxycholate may have survived this enrichment procedure (as was apparently true for DNA).

Figure 8

Demonstration of the operation of EtOH precipitates supporting transformation (experimental design 1.5; see Ref. [77]). A suspension of pneumococci that had been collected by centrifugation was heat-killed and extracted in the presence of deoxycholate. After removal of the cells by centrifugation, the extract was precipitated by increasing concentrations of ethanol in combination with Ca²⁺ (Sevag method, see Ref. [78]). The redissolved pellet fractions were dialyzed, then displayed an either fibrous (25% ethanol, final conc.) or flocculent (66% ethanol, final conc.) appearance. The two distinct precipitates were tested for the presence of both LPS of serotype S (III) or nucleic acids and transforming activity before stockpiling by freeze drying. PM vesicles, as well as micelle-like LP/LPS complexes, were certainly eliminated from the water-clear factor in the course of the multi-stage preparation of the whole-cell homogenate, which included detergent solubilisation and subsequent fractionated precipitation with ethanol. Detergent-solubilized LPS and LP may have been eliminated from the transforming principle in the course of fractionated ethanol precipitation (as was apparently not true for DNA).

Figure 9

Exclusion of S (III) LPS and RNA as the transforming principle (experimental design 1.6.1; see Ref. [80]). Upper section: An ethanol-precipitated, freeze-dried, redissolved preparation of the S (III) LPS was treated with Dubos SIII enzyme [81]. The digested materials did not display any serologically detectable S (III) antigen and, concomitantly, did not affect the transforming activity of the enriched preparation of S (III) LPS, which, per se, turned out to account only for slightly above background levels. Lower section: Pneumococci grown at low initial inoculum and glucose levels were recovered by centrifugation and, after resuspension, subjected to heat killing. Following centrifugation, the pelleted cells were washed with salt solution and re-centrifuged. This cell pellet was submitted to extraction by deoxycholate and re-centrifuged again.

Figure 10

Pneumococci grown at low initial inoculum and glucose levels were recovered by centrifugation and, after resuspension, subjected to deproteinization (experimental design 1.6.2; see Ref. [80]). Following redissolvation, the deproteinized materials were digested by both SIII Dubos enzyme [81] and ribonuclease, followed by removal of the added enzymes by repeating the Sevag method [78] and of the cleavage products of the digestion by dialysis. After the addition of increasing volumes of ethanol, precipitates of different characteristics became visible in sequential fashion from 50% to 66% ethanol (final conc.)—first as fibrous precipitates and later as more flocculent precipitates. Strikingly, the major portion of the total transforming activity was recovered with the fibrous fraction.

Figure 11

Comparative analysis of mammalian tissue and pneumococcal “chromosin” (experimental design 1.7; see Refs. [80,82]). Upper section: Organs like the liver, spleen, and thymus were blended repeatedly with physiological salt solution, which dissolved away most of the soluble materials in the case of application to pneumococci but left behind an insoluble remnant constituted largely by cell nuclei upon use for mammalian tissue lysis. The nuclear fraction was suspended in salt solution harbouring seven times the concentration of physiological salt solution. This procedure immediately resulted in the formation of a viscous solution of nucleoprotein, which was precipitated as fibrous strands upon repeated redissolvation in strong salt solution and reprecipitation by dilution. The attraction between the (very acidic) DNA and the (very basic) histones was counteracted by high concentrations of sodium and chloride ions, which necessitated the presence of levels of high salt for dissolvation of mammalian nucleoproteins. To obtain pure DNA from this mixture, Mirsky applied the same Sevag method of deproteinization by chloroform, which yielded a clear solution of DNA, irrespective of the salt concentration. In the course of precipitation of the DNA by pouring a thin stream of its solution into EtOH and constant stirring, fibrous precipitates, i.e., white bundles of DNA fibres, wound around the stirring rod, so they could be simply collected and lifted out as a single mass around the rod. Lower section: Pneumococci were grown at low glucose levels; sequentially deproteinized by EtOH precipitation; digested with Dubos SIII enzymes and RNase, extracted with chloroform; and, finally, precipitated with EtOH. After washing of the pellet with ether and its drying, the resuspended materials were poured into alcohol, leading to fibrous precipitates, which were wrapped around the stirring rod. Alternatively, some of the freeze-dried pneumococcal preparations were poured into EtOH, resulting in the formation of fibrous precipitates that were wrapped around the stirring rod and, in addition to similar fibres that floated in the solution. Most importantly, more than half of the DNA reacted with diphenylamine, and simultaneously, more than half of the transforming activity was recovered with the fraction that was wrapped around the rod. In contrast, the major portion of the RNA was left with the materials that had not been collected by the rod.

Figure 12

Analysis of the transforming principle using ultracentrifugation (experimental design 1.8; see Ref. [80]). Pneumococci grown at low glucose levels were sequentially fractionated in multiple precipitation, chloroform extraction and freeze-drying procedures. The fractions liberated from both protein and S (III) LPS, then redissolved in H₂O, were subjected to preparative ultracentrifugation (30,000 rpm, 4 h). At the bottom of each of the centrifuge tubes, a gelatinous pellet was recovered after pouring off the supernatant. Upon redissolvation in 140 mM NaCl, the pellet fraction was demonstrated to display most of the transforming activity, as well as the major portion of the DNA content of the starting fraction. In contrast, most of the RNA, S (III) LPS and additional pneumococcal antigens were recovered with the supernatant.

Figure 13

Biochemical evidence of the DNA nature of the transforming principle (experimental design 1.9; see Refs. [83,84,85,86,87,88,89,90]). I. (i) Crude enzyme preparations from various mammalian sources were incubated with crude pneumococcal fractions, as well as with DNA, which was partially purified from calf thymus. Each of the preparations with the potential to destroy the transforming activity also managed to degrade the calf thymus DNA. I. (ii) Dog, human and rabbit sera caused inactivation of the transforming activity, as well as cleavage of calf thymus DNA, which were abrogated upon heating of the sera to 60 °C and 65 °C, respectively. Thus, the inactivation pattern of the DNA-cleaving activity of the sera exactly matched that of the disruption of the transforming activity. II. S (III) pneumococci were heat-killed (65 °C, 30 min); precipitated with EtOH; washed with 140 mM NaCl; extracted for the nucleoprotein; and, finally, precipitated by EtOH. After repeated cycles of redissolving and reprecipitating, followed by extraction with deoxycholate and final precipitation with EtOH, a considerable amount of material was recovered, which harboured the expected transforming activity. III. Analysis of composition and purity by a variety of methods revealed that this material was constituted by deoxyribonucleoprotein very similar to Mirsky’s mammalian “chromosins” [82,86].

Figure 14

Transfer of DNA and, putatively, other matter of inheritance between donor and acceptor bacteria (A–D; adapted with modifications and permission from Ref. [115], 1973, Fritz Kaudewitz, Springer Press). (A) Electron microscopy of F⁻ acceptor and Hfr strains of donor cells of Escherichia coli K-12, as defined by low and high rates of recombination, respectively, connected by a long tube-like sexual pilus originating from the latter. The donor bacterium was decorated with appropriate fili-like RNA phages adsorbed by the cell body. For the discrimination from the E. coli-typical elongated thin shape, mutant acceptor bacteria exhibiting a shortened, stocky cell shape were used [116]. (B) Interpretation of the mechanism of the “rolling-circle” replication of DNA in the course of chromosome transfer from Hfr to F⁻ cells, which anticipates the transfer of solely the (+) strand following copying of the intact (−) strand as the template in the Hfr cell (broken line) before initiation of the transfer. In the F⁻ cell, the transferred (+) strand is supplemented with the complementary (−) strand (broken line), resulting in a DNA double helix. (C) Electron microscopy of a pair of Hfr and F⁻ E. coli cells (for their discrimination, see Ref. [117]). (D) Interpretation of the mechanism of transfer of DNA from E. coli K-12 Hfr to F⁻ cells, which anticipates the building-up of a bridge consisting of a pilus-like PM structure filled with cytoplasm connecting them.

Figure 15

Infection of bacteria by bacteriophages for the demonstration of DNA as the matter of inheritance. (A) Electron microscopy (adapted with permission from Ref. [115], 1973, Fritz Kaudewitz, Springer Press). Upper section: Replication of bacteriophage Phagus lacticola, exhibiting a very thin and long tail structure in cells of Mycobacterium spec. A number of phage particles was already adsorbed by the cell. After successful injection of the phage DNA, the deeply buried phage, consisting only of the proteinaceous capsid, became easily penetrated by electrons due to the missing DNA (grey colour; see (B), stage iv vs. the electron-dense “white” DNA-filled capsids; see (B), stages i–iii). The cell in the upper-left section was filled with newly assembled phage particles due to the high phage multiplicity applied and was already prepared to undergo lysis. This led to the release of infectious phage particles. Middle section: Electron microscopy of particles of phage T4 following adsorption by E. coli cells at high multiplicity. The previously extended tail sheaths were already contracted (see (B), stages i/ii vs. iii/iv). Lower section: Electron microscopy of the ultra-thin section of the momentum of injection of phage DNA across the cell wall (see (B), stage iv), which was penetrated by the tail tube (see (B), stages iii and iv) upon first contact of the tail fibres (see (B), stage i) and tail tip (see (B), stage ii; recognizable if in the plane of section) with the cell surface (site of injection indicated (see arrow in (A)). The previously extended tail sheaths were already contracted. The injected DNA was clearly visible (see (B), stage iv). (B) Interpretation, including hypothetical explanations (see text for details). (C) Hershey–Chase experiment for demonstration of independent functions of the capsid polypeptides and the DNA of bacteriophage particles (adapted with modifications and permission from Ref. [115], 1973, Fritz Kaudewitz, Springer Press) (see text for details).

Figure 16

Flow chart of the scientific reductionism leading from cellular to DNA heredity via topological and macromolecular heredity (green arrows) under the accompanying sequential elimination of “cybernetic” and “structural” information in the course of cell division (red arrows), including regulatory circuits and cellular membranes, and transformation (red arrows), including OM/PM-vesicles and proteins, under concomitant narrowing of the capability of protein synthesis by the elimination of non-DNA matter (red arrow). The biogenesis of macromolecules, cellular topology and total cells critically depend on folding, self-assembly and self-organization, respectively, as indicated. The levels of the various designs of the Griffith transformation experiment (see Ref. [21]) and its successors (see Refs. [22,70,71,72,73,74,75,76,77,91,92,93,94,95,96,112,113,114,118]) on the neglect, exclusion and emphasis of the transfer of information for cybernetics, structures and protein synthesis, respectively, are indicated.

Figure 17

Putative role of structural inheritance for the transition of donor R to acceptor S pneumococci in the Griffith transformation experiment (design 1.0). (A) Donor pneumococci of serotype S display a (lipo-)polysaccharide capsule consisting of LP and LPS, which are encoded by corresponding genes (pink circles) embedded in the total bacterial genome (green circles). (a) Their growth was attenuated by the presence of chocolate broth and anti-S antiserum raised against serotype S. (b) Attenuated S pneumococci were co-injected, together with acceptor pneumococci of serotype R, into mice. The attenuation procedures led to increased susceptibility of S pneumococci to (limited) degradation and lysis, which are caused by the specific conditions of the host organism and result in the release of DNA fragments (green colour, genome encoding the R phenotype, i.e., lacking the gene responsible for LPS synthesis; pink colour, gene responsible for LPS synthesis), micelle-like LP/LPS complexes and PM vesicles with LP and LPS inserted into their outer leaflet. (c) It remains to be clarified whether the anti-S antiserum, which becomes diluted in the course of the injection, “actively” participates in the production of the micelle-like LP/LPS complexes and/or PM vesicles. (d) The released DNA fragments encoding genes for the synthesis of the S-antigen (pink circles), as well as the LP and/or LPS, are integrated into the genome of acceptor pneumococci of the R phenotype (green circles) by recombination, insertion or fusion, respectively. It critically depends on the expression of pre-existing DNA and PM with their LP and LPS constituents. The latter structures operate as a template to enable mutual interactions and the correct assembly of the newly synthesized and subsequently inserted (micelle-like LP/LPS complexes) or fused matter (PM vesicles) into the capsule of serotype S. (B) Pneumococci of different species, e.g., Streptococcus agalactiae, do not harbour LP and/or LPS (green spheres) in their PM throughout their “history” but unrelated “false” counterparts (turquoise cuboids), as well as DNA not related to the genome of Streptococcus pneumoniae (turquoise circles). (e) They are not able to acquire, replicate and transfer the new virulent S serotype.