Aspergillus fumigatus is a ubiquitous environmental mold. When conidia of A. fumigatus are inhaled by immunosuppressed individuals they can germinate to form filamentous hyphae that invade lung tissue to cause a necrotizing pneumonia which is associated with a high mortality rate. One of the major virulence factors of A. fumigatus is the production of the secreted cationic exopolysaccharide galactosaminogalactan (GAG), which is essential for biofilm formation. GAG is predicted to be synthesized by the products of a five gene cluster, one of which, ega3, is predicted to encode a glycoside hydrolase anchored to the cell membrane of A. fumigatus.Through mass spectrometry studies, we established that Ega3 cleaves mature deacetylated GAG. We hypothesized that like the other proteins in the cluster, Ega3 is necessary for GAG synthesis. To test this hypothesis, we sought to disrupt the ega3 by allele replacement with a drug resistance marker and characterize the phenotype of the resulting ega3 null mutant. Initial attempts to disrupt ega3 by our usual approach were unsuccessful but two ega3 null mutants were finally recovered, one using CRISPR/Cas9 and another using the conventional split marker protocol. Like other mutants in the GAG cluster, these mutants were deficient in biofilm formation and did not produce deacetylated GAG. Surprisingly, complementation with an ega3 allele failed to restore biofilm formation in both strains, despite restoration of Ega3 protein production as demonstrated by Western blot. These findings were consistent with the presence of a secondary mutation impairing GAG production. Analysis of the expression of the GAG cluster genes revealed that agd3, encoding the GAG deacetylase required for the production of mature, cationic GAG, was not expressed in the ∆ega3CRISPR mutant. Genome sequencing revealed that uge3, required for GAG production, was mutated in the other ega3 null mutant. Since we were only able to disrupt ega3 in the absence of the production of cationic GAG, we hypothesized that ega3 is conditionally essential in the presence of cationic GAG. To test this hypothesis, agd3 was expressed in the ∆ega3CRISPR mutant under the control of a tetracycline-inducible promoter (∆ega3CRISPR::agd3Tet on). Under agd3- expressing conditions, GAG production was restored but fungal growth was inhibited.Since GAG is a large cationic polymer and Ega3 cleaves cationic GAG, we hypothesized that cationic GAG may be toxic to the cell membrane of A. fumigatus and that membrane-bound Ega3 degrades GAG near the membrane. To test whether cationic GAG induces damage to the cell membrane in the absence of Ega3, ATP was measured in fungal culture supernatants as a proxy for cell leakage. Induction of agd3 expression in the (∆ega3CRISPR::agd3Teton) mutant resulted in the release of high levels of ATP suggesting that cationic GAG disrupts the cell membrane of A. fumigatus. We therefore hypothesized that secreted GAG may also mediate host cell injury during infection. To test this hypothesis, A549 pulmonary epithelial cells were loaded with radioactive chromium and exposed to culture supernatants from wild- type A. fumigatus+/- Ega3. Exposure to GAG-containing culture supernatants (CS) induced epithelial cell injury, but cells were almost completely protected from damage in the presence of Ega3. Propidium iodide (PI) staining of BMDMs exposed to CS mirrored these results, as did PI/annexin V staining of human NK cells.