Spongia officinalis has been used for millennia and for various purposes: from the Greeks for bathing and lining their armours, to today's pharmaceutical attempts to produce anti-inflammatory and other therapeutics, not even mentioning the use by Arabic physicians as early as 932 A.D of soaked sponges with narcotic drugs to place over the patient’s nose to provide a state of anesthesia. In a nutshell everyone knows the "bath sponge".

Spongia officinalis is a leuconoid sponge. A leuconoid sponge has a thick body wall, and the ostia open into incurrent canals that draw water into the sponge’s body. These incurrent canals open into chambers that are lined with choanocytes. Water flows from these chambers into excurrent canals that empty into a relatively small spongocoel. From there, water exits through an osculum.

Unlike asconoid and syconoid sponges, which are basically built around their spongocoels and oscula, a leuconoid sponge has a complex, irregularly-shaped body that may have several oscula. It allows a leuconoid sponge to grow to larger size.

It also provides them with powerful filtration capacity (approximately 1200 times its volume per day) which makes it a perfect host for our system. But as mentioned earlier our aim is to engineer an epibiont and not the sponge itself.

Following this introduction, we will both describe what Pseudovibrio denitrificans exactly is and how we turned it into a transformable/selectable-ready chassis.

At first glance, working on the microbiome of sponges appears like a hazardous task: 5734 articles for marine sponges, only 51 with microbiome of sponges, only 28 articles mentioning Pseudovibrio genius, only 5 mentioning Pseudovibrio denitrificans, and finally none about a genetic engineering system in it.

Moreover with only 12 species in the genius, and 2 strains sequenced (Pseudovibrio sp. FO-BEG1, Pseudovibrio sp. JE062) we knew we would have to sequence our strain of Pseudovibrio denitrificans with a mapping to one of the 2 strains references. A formidable drawback even considering that in the first place the species was chosen because of its natural denitrification ability (see toxic compound) and easy to source at DMSZ.

On the other hand Pseudovibrio denitrificans, or strains related to the species, has been shown to be in the majority in at least six microbiomes of sponges in the Mediterranean Sea, where spongia officinalis resides.

Now the thorny question is: can it be easily used in our lab?

With the help of the 'Molecules of defense and communication in the microbial ecosystems' team of the National museum of natural history of Paris (their web page here), we succeed to cultivate Pseudovibrio denitrificans and the strain Pseudovibrio ascidiaceicola (that they have generously given to us) in two medium, the marine broth (MB) and the minimal medium M9 added 2% of CASAmino Acid and by 3% of NaCl.

Table1: Composition of used media, Marine Broth and M9-CASA+3% NaCl.

Because of the opaqueness of the marine broth, it was impossible to obtain reliable data of the growth of the bacteria. We try to have a curve of growth in filtered MB but even if the bacteria grow, it's not the same condition. Thus, the growth curve has been made in M9-CASA+3%NaCl 1X, with the TECAN infiniteM200 (cf: Figure1).

Figure1: Curve of growth in liquid M9-CASA+3%NaCl 1X and liquid MB, and microscopic photo of Pseudovibrio denitrificans (left) and Pseudovibrio ascidiaceicola (right).

The growth of our bacteria in 200µL of M9-CASA+NaCl is not very good, moreover the ratio culture/air is not optimal in the TECAN wells. In bigger cultures, the growth is faster.
The marine broth is more complete than the M9-CASA+3%NaCl, even if we have adjusted the salinity. As we can see in the Table 2, Pseudovibrio denitrificans grows better on MB medium. The number of colony is higher, and they are significantly bigger (cf: Table2). Moreover, the growth is faster.

Table2: Number of CFU of MB 1X and M9-CASA+3%NaCl 1X plates.
The number of CFU was calculated by inoculated respectively 10µL of pre-cultures of Pseudovibrio denitrificans in MB 1X and in M9-CASA+3%NaCl at OD (600nm) ~ 1.

In the aim to know if we can transfer Pseudovibrio denitrificans from a media, to the other one, we have made some tests and calculate a ratio of survivability. We have made two pre-cultures of Pseudovibrio denitrificans in MB 1X and M9-CASA+3%NaCl 1X, and after a culture overnight, we streaked them on plates of MB 1X and M9-CASA+3%NaCl 1X. The CFU on each plate were counted and ratios of survivability were calculated using the formula Sample/Control (controls being the number of CFU obtained on a plate from a pre-culture in the same medium). (cf: Figure2).

Figure2: Tests of survivability, table of ratio and photos (on the left: M9 plates, on the right: MB plates).

Pseudovibrio denitrificans can survive even if we transfer them from a medium to the other one. But we can see that even cells come from a pre-culture of M9-CASA+3%NaCl, they grow better and faster on MB plates. Moreover, even if they come from a MB pre-culture, they don't grow very well on M9-CASA+3%NaCl plates.

First, we tried to transform Pseudovibrio denitrificans by a heat shock. Chemically competent cells were prepared with CaCl2 solution. Transformation protocols were carried out by mixing 50 µL of cells with 1 µL of plasmid pSB1C3 solution (35ng/µL). After the standard heat shock protocol for E.coli, cells were spread on 1X MB medium plates. We did not obtain any transformant from this protocol.
By studying transformation of bacteria phylogenetically close of our strain Pseudovibrio denitrificans, we noticed that marines bacteria are often resistant to common chemical transformation approaches (Piekarski T, Buchholz I, 2009) .

We considered testing protocols of conjugation and electroporation to transform our bacteria. Unfortunately, due to the lack of time we just tested the electroporation approach.
Two important parameters had to be finalized though; the type and the number of cell washes and also the voltage.

Washing test
Marine bacteria live in a salty environment. Thus, a classic water wash may to lead to an osmotic shock. Therefore, we tested washes with glycerol 10% and sorbitol 2% to try to limit the shock.

Figure6: Protocol of washing tests

Pre-cultures of Pseudovibrio denitrificans and E.coli cells were prepared and grew in an shaking incubator overnight at 30°C and 37°C respectively. After performing the protocol (see section Protocols) of electrocompetents cells, 20µL of pre-cultures were spread when the DO(600nm) reached 1.5 for Pseudovibrio and 0.5 for E.coli, on plates corresponding to the media used for pre-cultures. CFU were then counted for all plates (Table5).

Table5: Results of washing test, number of CFU for the different washes.

Washes tests reveal that sorbitol is the best way to wash cells in MB pre-culture. Gycerol can also be used. For pre-cultures in M9, the better way to wash cells is by using glycerol.

Voltage test
For this part, we had to find the voltage which allows the best efficiency, by not killing the cells but by being important enough to transform our bacteria.
To achieve this goal, we developed a similar protocol as the standard protocol described previously, by testing different voltages on our cells, without plasmids (Figure7).

Figure7: Protocol of voltage tests

Pre-cultures of Pseudovibrio denitrificans and E.coli were prepared and grew overnight in a shaking incubator, at 30°C and 37°C respectively. After performing the protocol (see section Protocols) of electrocompetents cells, we electroporated our cells at either 1200 V, 1800 V, 2000 V and 2200 V. Cells were then incubated respectively at 30°C for Pseudovibrio and 37°C for E.coli. Then we plated the cells on MB or M9 plates according to the media used for the pre-culture.

Unfortunately, some plates were convered by a cellular lawn and it was impossible to count the CFU. We chose to estimate roughly the cell growths as shown in the Table 6.

Table 6: Results of voltage tests.

In this table we can see that the higher voltage for which we obtained a very good rate of living cells is 2000 V for Pseudovibrio. According to these results, we chose to do a 2000 V electroporation to transform our Pseudovibrio denitrificans

Test with pSB1C3
This year, all the biobricks had to be sent in the backbone pSB1C3. For our first try to transform Pseudovibrio denitrificans, we thus chose to test this plasmid.
We made competent cells and then we did the developed transformation protocol (see section Protocols), which is an electroporation at 2000 V.

Figure8: Maps of PSB1C3.

After a lot of tries, we did not succeed to have transformant of Pseudovibrio denitrificans which could grow on plates supplemented with chloramphenicol.
Because this plasmid works in E.coli which is a GRAM- bacteria as Pseudovibrio denitrificans, we expected it will work in our strain as well. Unfortunately, the plasmid was not expressed.

Other plasmids tested
After trying transforming our strain with a standard plasmid of iGEM, pSB1C3, we tried other plasmids, pBBR1MCS (Katzke N, Arvani S, 2010) and pRhokHI-2 (Katzke N, Bergmann R, 2012). Those plasmids were generously given by Dr Thomas DREPPER from the Heinrich-Heine-Universität Düsseldorf, Institute of Molecular Enzyme Technology (Group of Bacterial Photobiotechnology). Dr DREPPER succeed to transform several strains of a genus very closed to our Pseudovibrio, the genus Roseobacter.
We made competent cells and then we effected the developed transformation protocol (see section Protocols), which is an electroporation at 2000 V.

Figure9: Maps of PBBR1MCS and pRhoKHI-2.

Once again, after a lot of tries, we failed to have transformant of Pseudovibrio denitrificans growing on plates supplemented with chloramphenicol, even if pBBR1MCS is a plasmid supposed to work in almost all GRAM- bacteria and those two plasmids work in bacteria very close to Pseudovibrio.

Creating a new plasmid
After trying to transform Pseudovibrio denitrificans with the pSB1C3 plasmid which did not work, we had several hypothesis:
-The promoter of the Chloramphenicol resistance gene does not work in our bacteria
-The origin of replication does not work in our bacteria
-There is a system degrading exogene DNA in Pseudovibrio
To verify these hypothesis, we chose to find specific constitutive promoters and origin of replications in an other member of the genus Pseudovibrio, which genome has been sequenced, the FOBEG1 strain. (lien article)

A strong constitutive promoter of FOBEG1 was found with a genome browser. Upstream sequences of vital and cell cycle independant genes were explored. We were interested by the transkelotase, a key enzyme of the pathway of pentoses phosphates.
We can assume this enzyme should have a good constitutive promoter, or at least have a reliable and strong expression. To be sure to have the whole promoter sequence, we exported the sequence (fasta format) from the beginning of the ORF of the transkelotase, to the end of the previous ORF. This sequence was amplified with primers 5 and 6 (see primers table in Protocols).

After a PCR, we obtained a 6000pb band for Pseudovibrio ascidiaceicola, and no band for Pseuvibrio denitrificans. These PCR product was sent to sequencing, as shown in figure M.

Figure10: Result of promoter sequencing.

Thus, there is 94.1% of identity between the transketolase promoter of the Pseudovibrio FOBEG1 strain and the Pseudovibrio ascidiaceicola strain.

For the replication origin, the aim was here to find the replication origin of the plasmid of the reference strain FOBEG1. To do so, we looked for the repA, repB and repC genes on NCBI. Those three genes follow each other in the genome.
This sequence was amplified with primers 49 and 50 (see primers table in Protocols).

Figure11: Result of ORI amplification in Pseudovibrio denitrificans (Pd), Pseudovribrio ascidiaceicola (Pa) and E.coli DH5a.

The negative control with E.coli DH5a confirmed that there is no amplification. For Pseudovibrio strains, we obtained one band for Pseudovibrio denitrificans and two for Pseudovibrio ascidiaceicola.


Meanwhile the plasmid of the transposase was tested, and the project to create a viable plasmid for pseudovibrio was abandoned (Transposons ).
Moreover, our third hypothesis of a system which would degrade exogen DNA was supported by BLAST data, and the presence of the enzyme EcoRI in the genome (see section Genome assembly). This enzyne is indeed known to degrade DNA without a certain pattern of methylation.