Table 1.

Summary of animal models of paroxysmal nocturnal hemoglobinuria.

Reference	Model	Tissue-specificity/Cre	Selected Observations
Germline Piga loss mouse model
Kawagoe et al. (1996) (32)	Constitutional Piga knockout; mice chimeric for Piga deficiency.	n/a	Low chimerism in non-hematopoietic tissues due to embryonic lethality in mice with high proportion of GPI (-) cells. Percentage of GPI (-) hematopoietic cells did not increase over time.
Rosti et al. (1997) (31)	Constitutional Piga knockout; mice chimeric for Piga deficiency	n/a	In mice chimeric for Piga deficiency, Piga embryonic stem cells were competent for trilineage hematopoiesis; there was no expansion of GPI (-) cells.
Chimeric mouse model
Tremml et al. (1999) (33) Keller et al. (1999) (34)	Constitutional Piga knockout; mice chimeric for Piga deficiency	EIIa-Cre; early embryonic	Mosaic mice are viable. GPI (-) RBCs have increased sensitivity to complement mediated lysis and shorter half-life; comparatively high fraction of GPI (-) lymphocytes; no clonal expansion of GPI (-) cells overtime.
Hematopoietic-specific PIGA loss mouse model
Keller et al. (2001) (35)	Hematopoietic-cell-restricted knockout of Piga	Fes-Cre; all hematopoietic cells	GPI (-) cells can fully reconstitute trilineage hematopoiesis, have differences in lineage contribution (e.g., more GPI (-) T cells), but have no clonal expansion overtime.
Jasinski et al. (2001) (36)	Erythroid/megakaryocyte- restricted knockout of Piga	GATA1-Cre; erythroid-megakaryocyte lineage	Leaky expression in early embryogenesis leading to high embryonal lethality. Mice that escaped embryonal recombination had almost 100% of red cells with partial deficiency of GPI-anchored proteins, and intermediate sensitivity to complement, resembling type II PNH cells.
Visconte et al. (2010) (37)	Hematopoietic-cell-restricted knockout of Piga	Fes-Cre; all hematopoietic cells	No hemolysis; the frequency of GPI (-) cells was much higher in T lymphocytes but lower in erythrocytes, granulocytes, and B cells.
Hazenbos et al. (2004) (38)	T lymphocyte-restricted knockout of Piga	Lck-Cre; T lymphocytes	Stimulation by ConA or Allogeneic stimulation of GPI (-) T cells induced higher proliferative responses than normal cells.
Hazenbos et al. (2011) (39)	Hematopoietic-cell-restricted knockout of Piga	Vav-Cre; all hematopoietic cells	Grossly normal numbers of T and B lymphocytes, monocytes, neutrophils, and erythrocytes; detailed hematopoietic analysis not performed.
Fetal liver transplant mouse model
Murakami et al. (1999) (40) Murakami et al. (2002) (41)	A transplant model of hematopoietic cell-restricted knockout of Piga	CMV-Cre, followed by transplantation of fetal liver cells from female mice mosaic for Piga-deficiency into wild type recipients.	No expansion of GPI (-) cells overtime. Frequency of GPI (-) cells was highest in T lymphocytes and immature thymocytes. GPI (-)cells engrafted following transplantation, but did not outcompete wild type cells. When transplanted with CD4+ allo-reactive to donor hematopoietic cells, GPI (-) donor cells were less sensitive to CD4+ T cell-driven immune attack.
GPI-anchored complement regulator knockout mouse model
Sun et al. (1999) (42)	GPI-DAF-Deficient Mice	n/a	No embryonic lethality; no hemolytic anemia
Lin et al. (2001) (43)	Daf-1 knockout mice	excision in ESCs	No hemolytic anemia; increased C3 deposition on erythrocytes of Daf-1 -/- mice.
Holt et al. (2001) (44)	Cd59a knockout mice	n/a	No hemolytic anemia; elevated reticulocytes; erythrocytes susceptible to complement lysis.
Miwa et al. (2002) (45)	Cd59a knockout mice and Cd59a/Daf double knockout mice	n/a	High sensitivity to complement lysis but no spontaneous hemolytic anemia.
Qin et al. (2003) (46)	Cd59b knockout mice	n/a	Spontaneous hemolytic anemia with morphological abnormalities in RBCs and platelets.
Non-human primate model
Shin et al. (2019) (47)	CRISPR/Cas9 PIGA gene editing in Rhesus Macaque	n/a	No clonal expansion of GPI (-) cells.

n/a, not applicable.