Research program.

During my doctorate with Katy C. Kao at Texas A&M University I worked on evolutionary engineering, which uses evolution in the laboratory to obtain microorganisms with new traits. In Escherichia coli we found 14 genes whose modification raises tolerance to n-butanol by up to 49 %. In yeast, evolution produced mutants that grew up to 57 % faster in biomass hydrolysates. With hydrogen peroxide as a selective pressure, we tripled the carotenoid production of an engineered yeast, to 18 mg per gram of dry biomass, in what is still my most cited article.

In 2014, at the National Renewable Energy Laboratory (NREL) in the United States, we engineered Clostridium thermocellum, a bacterium that breaks down cellulose, to also consume xylose, the sugar of hemicellulose. With both sugars it produced twice as much hydrogen and ethanol as with cellulose alone. In 2020, already at Uniandes, Albert E. Tafur Rangel and I identified the enzyme that limited its growth on xylose. In 2024, with adaptive evolution and an added enzyme, we obtained the first strain described that assimilates hemicellulose without losing its ability to break down cellulose. That strain is the subject of a United States patent application on which I am a co-inventor.

At Uniandes, this foundation became the line of reverse engineering of microorganisms, which studies how evolution solves the problems of a microbial factory.

A drug, a gene or a protein is of little use if it does not get inside the cell. It has to cross the membrane and then leave the endosome, the compartment where the cell breaks down almost everything it takes in. With Juan Carlos Cruz, of Biomedical Engineering, we design nanoparticles for that journey: cores of magnetite or sheets of carbon, coated with peptides that cross membranes, such as buforin II. Carlos Manuel Ramírez Acosta, my master’s student, made one of the first: a magnetite core with a silver shell and a polymer that releases DNA when the pH drops inside the cell. The magnetoliposomes of the therapies for lysosomal diseases grew out of it.

Julian Daniel Torres Vanegas, whom I advised in his doctorate, studied what happens when the support and the peptide are changed. With graphene oxide and buforin II, the particles covered close to 100 % of the cells in less than 30 minutes; without the peptide, between 40 and 70 %. In Vero cells, between 30 and 45 % escaped from the endosome. His review of the barriers these vehicles face is the most cited of those I have published from Uniandes. Laura Tatiana Salgado, whom I co-advised in her master’s, tested where in the peptide to place the motif that opens the way through the membrane. With the motif at the free end, the particles entered more and escaped the endosome better than those of the other two designs and those carrying known peptides such as buforin II.

We then took these vehicles to specific diseases. Erika Alejandra Díaz Ramírez, whom I co-advised in her master’s, loaded magnetite nanoparticles with temozolomide and paclitaxel, two cancer drugs. They were more toxic to melanoma cells than to skin keratinocytes, and outperformed the free drugs used at concentrations four times higher. David Arango, whom I also co-advised, built in his master’s a magnetite platform to carry gene editing tools to neurodegenerative diseases. In the Parkinson’s disease project led by Natasha Bloch, where I was a co-investigator, magnetite nanoparticles with buforin II raised the expression of the neuroprotective gene gdnf up to 214-fold in a cell model of the disease, against 41-fold with lipofectamine.

Cristian F. Rodríguez, whom I co-advised in his master’s and who is now an instructor professor at Uniandes, put magnetic nanoparticles inside spheroids of cells. With an external magnetic field, two spheroids fused in about 2 days instead of about 7, which speeds up the fabrication of tissues in the laboratory. He also wrote our review of magnetoliposomes (2025). In 2024 Juan Carlos Cruz and I were guest editors of a collection of Nanoscale Advances, of the Royal Society of Chemistry, on nanoplatforms that respond to stimuli, and in 2025 I co-edited with him and Yashwant V. Pathak the book Nanocarriers for Nucleic Acids and Proteins (CRC Press).

Many healthier or more sustainable foods lose out to conventional ones because of what consumers expect from them and because of their texture. We asked whether both can be designed. With Felipe Reinoso-Carvalho, of the School of Management, I advised the doctorate of Brayan Rodríguez, now a professor at Pontificia Universidad Javeriana. In experiments with 592 people, the sound of a beer poured from the bottle reduced negative emotions toward non-alcoholic beer and raised its expected freshness. In work with Felipe’s collaborators at Instituto Universitário de Lisboa (Iscte), sounds of nature strengthened the perceived environmental sustainability of a product, and sounds of groups of people its social sustainability, in two studies with 517 participants.

In the last chapter of his doctorate, Brayan trained a deep learning model, on music that people had already rated, to predict how sweet, bitter, salty, sour or spicy a piece of music sounds. With those predictions we built Sonic Seasoning, an app that assembles playlists for the flavor profile and the genre you choose. The article is under review; the app is public and anyone can use it.

We study texture with Martha Liliana Díaz-Bustamante, now my doctoral student, and with Óscar Alberto Álvarez Solano. Two costeño cheeses with half and a quarter of the salt, cooked at 55 °C, resembled the control in composition and texture. In a soy yogurt, replacing 25 % of the protein with yeast grown on whey raised firmness by 54 % and cut fermentation to seven or eight hours, against almost 17. How consumers receive it is still to be measured.

Yeast shapes much of the flavor and aroma of a beer, and Carlos Yesiel González Murcia looks for it in Colombian biodiversity. Since 2026 he has been doing a Doctorate in Engineering on native yeasts for brewing, advised by Nicolás Ríos Ratkovich, of my department, and co-advised by me.